Joseph M. Pierre, MD

Many psychiatrists conceptualize mental illnesses, including psychotic disorders, across a continuum where their borders can be ambiguous.¹ The same can be said of individual symptoms such as delusions, where the line separating clear-cut pathology from nonpathological or subclinical “delusion-like beliefs” is often blurred.^2,3 However, the categorical distinction between mental illness and normality is fundamental to diagnostic reliability and crucial to clinical decisions about whether and how to intervene.

Conspiracy theory beliefs are delusion-like beliefs that are commonly encountered within today’s political landscape. Surveys have consistently revealed that approximately one-half of the population believes in at least 1 conspiracy theory, highlighting the normality of such beliefs despite their potential outlandishness.⁴ Here are 3 questions you can ask to help differentiate conspiracy theory beliefs from delusions.

1. What is the evidence for the belief?

Drawing from Karl Jaspers’ conceptualization of delusions as “impossible” and “unshareable,” the DSM-5 distinguishes delusions from culturally-sanctioned shared beliefs such as religious creeds.³ Whereas delusions often arise out of anomalous subjective experiences, individuals who come to believe in conspiracytheories have typically sought explanations and found them from secondary sources, often on the internet.⁵ Despite the familiar term “conspiracy theorist,” most who believe in conspiracy theories aren’t so much theorizing as they are adopting counter-narratives based on assimilated information. Unlike delusions, conspiracy theory beliefs are learned, with the “evidence” to support them easily located online.

2. Is the belief self-referential?

The stereotypical unshareability of delusions often hinges upon their self-referential content. For example, while it is easy to find others who believe in the Second Coming, it would be much harder to convince others that you are the Second Coming. Unlike delusions, conspiracy theories are beliefs about the world and explanations of real-life events; their content is rarely, if ever, directly related to the believer.

Conspiracy theory beliefs involve a negation of authoritative accounts that is rooted in “epistemic mistrust” of authoritative sources of information.⁵ While conspiratorial mistrust has been compared with paranoia, with paranoia found to be associated with belief in conspiracy theories,⁶ epistemic mistrust encompasses a range of justified cultural mistrust, unwarranted mistrust based on racial prejudice, and subclinical paranoia typical of schizotypy. The more self-referential the underlying paranoia, the more likely an associated belief is to cross the boundary from conspiracy theory to delusion.⁷

3. Is there overlap?

Conspiracy theory beliefs and delusions are not mutually exclusive. “Gang stalking” offers a vexing example of paranoia that is part shared conspiracy theory, part idiosyncratic delusion.⁸ Reliably disentangling these components requires identifying the conspiracy theory component as a widely-shared belief about government surveillance, while carefully analyzing the self-referential component to determine credibility and potential delusionality.

Many psychiatrists conceptualize mental illnesses, including psychotic disorders, across a continuum where their borders can be ambiguous.¹ The same can be said of individual symptoms such as delusions, where the line separating clear-cut pathology from nonpathological or subclinical “delusion-like beliefs” is often blurred.^2,3 However, the categorical distinction between mental illness and normality is fundamental to diagnostic reliability and crucial to clinical decisions about whether and how to intervene.

Conspiracy theory beliefs are delusion-like beliefs that are commonly encountered within today’s political landscape. Surveys have consistently revealed that approximately one-half of the population believes in at least 1 conspiracy theory, highlighting the normality of such beliefs despite their potential outlandishness.⁴ Here are 3 questions you can ask to help differentiate conspiracy theory beliefs from delusions.

1. What is the evidence for the belief?

Drawing from Karl Jaspers’ conceptualization of delusions as “impossible” and “unshareable,” the DSM-5 distinguishes delusions from culturally-sanctioned shared beliefs such as religious creeds.³ Whereas delusions often arise out of anomalous subjective experiences, individuals who come to believe in conspiracytheories have typically sought explanations and found them from secondary sources, often on the internet.⁵ Despite the familiar term “conspiracy theorist,” most who believe in conspiracy theories aren’t so much theorizing as they are adopting counter-narratives based on assimilated information. Unlike delusions, conspiracy theory beliefs are learned, with the “evidence” to support them easily located online.

2. Is the belief self-referential?

The stereotypical unshareability of delusions often hinges upon their self-referential content. For example, while it is easy to find others who believe in the Second Coming, it would be much harder to convince others that you are the Second Coming. Unlike delusions, conspiracy theories are beliefs about the world and explanations of real-life events; their content is rarely, if ever, directly related to the believer.

Conspiracy theory beliefs involve a negation of authoritative accounts that is rooted in “epistemic mistrust” of authoritative sources of information.⁵ While conspiratorial mistrust has been compared with paranoia, with paranoia found to be associated with belief in conspiracy theories,⁶ epistemic mistrust encompasses a range of justified cultural mistrust, unwarranted mistrust based on racial prejudice, and subclinical paranoia typical of schizotypy. The more self-referential the underlying paranoia, the more likely an associated belief is to cross the boundary from conspiracy theory to delusion.⁷

3. Is there overlap?

Conspiracy theory beliefs and delusions are not mutually exclusive. “Gang stalking” offers a vexing example of paranoia that is part shared conspiracy theory, part idiosyncratic delusion.⁸ Reliably disentangling these components requires identifying the conspiracy theory component as a widely-shared belief about government surveillance, while carefully analyzing the self-referential component to determine credibility and potential delusionality.

Many psychiatrists conceptualize mental illnesses, including psychotic disorders, across a continuum where their borders can be ambiguous.¹ The same can be said of individual symptoms such as delusions, where the line separating clear-cut pathology from nonpathological or subclinical “delusion-like beliefs” is often blurred.^2,3 However, the categorical distinction between mental illness and normality is fundamental to diagnostic reliability and crucial to clinical decisions about whether and how to intervene.

Conspiracy theory beliefs are delusion-like beliefs that are commonly encountered within today’s political landscape. Surveys have consistently revealed that approximately one-half of the population believes in at least 1 conspiracy theory, highlighting the normality of such beliefs despite their potential outlandishness.⁴ Here are 3 questions you can ask to help differentiate conspiracy theory beliefs from delusions.

1. What is the evidence for the belief?

Drawing from Karl Jaspers’ conceptualization of delusions as “impossible” and “unshareable,” the DSM-5 distinguishes delusions from culturally-sanctioned shared beliefs such as religious creeds.³ Whereas delusions often arise out of anomalous subjective experiences, individuals who come to believe in conspiracytheories have typically sought explanations and found them from secondary sources, often on the internet.⁵ Despite the familiar term “conspiracy theorist,” most who believe in conspiracy theories aren’t so much theorizing as they are adopting counter-narratives based on assimilated information. Unlike delusions, conspiracy theory beliefs are learned, with the “evidence” to support them easily located online.

2. Is the belief self-referential?

The stereotypical unshareability of delusions often hinges upon their self-referential content. For example, while it is easy to find others who believe in the Second Coming, it would be much harder to convince others that you are the Second Coming. Unlike delusions, conspiracy theories are beliefs about the world and explanations of real-life events; their content is rarely, if ever, directly related to the believer.

Conspiracy theory beliefs involve a negation of authoritative accounts that is rooted in “epistemic mistrust” of authoritative sources of information.⁵ While conspiratorial mistrust has been compared with paranoia, with paranoia found to be associated with belief in conspiracy theories,⁶ epistemic mistrust encompasses a range of justified cultural mistrust, unwarranted mistrust based on racial prejudice, and subclinical paranoia typical of schizotypy. The more self-referential the underlying paranoia, the more likely an associated belief is to cross the boundary from conspiracy theory to delusion.⁷

3. Is there overlap?

Conspiracy theory beliefs and delusions are not mutually exclusive. “Gang stalking” offers a vexing example of paranoia that is part shared conspiracy theory, part idiosyncratic delusion.⁸ Reliably disentangling these components requires identifying the conspiracy theory component as a widely-shared belief about government surveillance, while carefully analyzing the self-referential component to determine credibility and potential delusionality.

When many people across the country, not to mention in public office, believe that the world is run by a group of Satanic pedophiles that includes top Democrats and Hollywood elites, and that former President Trump is leading a secret mission to bring these evildoers to justice, one can’t help but ask if they’re at least to some degree mentally impaired.

Conspiracy theories are often received with psychiatric connotations; associated with paranoid plan-hatchers, and nonbelieving outsiders. But whereas theories such as QAnon strain credibility for many people, we would argue that they are likely not the product of psychosis or mental illness; nor do conspiracy theories in general represent delusions.

For one thing, surveys have consistently revealed that about 50% of the population believes in at least one conspiracy theory. Furthermore, there are several substantive differences between conspiracy theory beliefs and delusions.

Some researchers consider conspiracy theories to be “a subset of false beliefs,” but  most scholars, ourselves included, do not prejudge their validity or veracity. Real-life conspiracies, such as the CIA’s MK-Ultra program, have clearly occurred throughout history.

Our central contention is that belief in conspiracy theories is distinct from psychosis, and more closely resembles extreme but subculturally sanctioned religious or political beliefs. However, the line between believing in conspiracies and being delusional becomes blurred when the believer becomes part of the conspiracy theory and feels compelled to act on the belief as part of a personal mission.

Take Edgar Maddison Welch, a 28-year-old man who firmly believed the so-called “Pizzagate” conspiracy theory – the baseless claim that Hillary Clinton and Democratic elites were running a child sex-trafficking ring out of a Washington, DC, pizzeria. Seeing himself a potential savior of children, Mr. Welch drove 350 miles to the pizza shop from his home in North Carolina in December 2016 and fired three shots from an AR-15 style rifle into a locked closet door, ultimately surrendering to police. However, on questioning he quickly conceded, “The intel on this wasn’t 100%.”

Who believes in conspiracy theories?

Given that half the population believes in at least one conspiracy theory, it should come as little surprise that there is no reliable “profile” for believers. Although some studies have suggested associations with low education, right-wing political orientation, and certain personality traits like subclinical paranoia and schizotypy, such findings have been inconsistent and may vary across specific conspiracy theory. Associations between conspiracy belief and paranoia suggest overlap within a “conspiratorial mindset,” with recent evidence that “distrust of officialdom” is a key mediator between believing in conspiracies and political ideology.

Other quantitative “cognitive quirks” reported in those who believe in conspiracies are a need for certainty and control, a need for uniqueness, illusory pattern perception, and lack of analytical thinking. It’s unclear which of these factors may represent universal cognitive explanations for conspiratorial beliefs, vs. those that might be related to specific beliefs, such as the need for certainty during times of crisis and societal upheaval, when conspiracy theories tend to flourish.

Much of the research on conspiracy theory belief is based on the questionable premise that it’s best understood at the level of the individual’s psychopathology, or the “deficit model,” as it’s called. One of us (JMP) has instead proposed a two-component model that includes social and informational contexts. The first component – epistemic mistrust – involves mistrusting conventional, “authoritative” knowledge. The second involves biased information processing and exposure to misinformation, often transmitted by word of mouth, or through social networks. With this model, believing in conspiracy theories could be conceived as involving “delusion-like beliefs,” but not frank psychosis or full-blown delusions, as one might see, for example, in schizophrenia.

Indeed, many of the cognitive characteristics associated with conspiracy theory belief are universal, continuously distributed traits, varying in quantity, rather than all-or-none variables or distinct symptoms of mental illness.

Essentially, delusions are fixed, false, usually unshared beliefs, often based on subjective “inner” experience. (One rare exception is the so-called folie à deux, in which two people appear to “share” the same delusion; however, psychiatrists have long debated whether both individuals should be considered truly delusional). The delusion’s content is often “self-referential”; i.e., focused primarily on the believer.

In contrast, conspiracy theories are usually, but not necessarily, false. They are typically shared beliefs that don’t explicitly or directly involve the believer, and are based on evidence that one finds “out there,” such as on the Internet. This speaks to the highly communal nature of so many conspiracy theories – networks of like-minded individuals reinforcing one another’s beliefs in a particular socio-cultural context.
 

Conspiracy theory belief, COVID-19, and medical intervention

As for medical conspiracy theories, none have flourished recently more so than those involving the COVID-19 pandemic. As a recent editorial by Stein and colleagues noted, “Some conspiratorial claims include assertions that COVID‐19 is a hoax; arguments that the virus was created artificially and spread on purpose as a bioweapon; or allegations that governments are using the emergency situation to pursue their antidemocratic goals. … Other conspiracies argued that people in power are taking advantage of the pandemic as a plan to inject microchip quantum-dot spy software and monitor people.”

Stein and colleagues make the important point that a “key difference between COVID‐19 and the 1918 flu pandemic ... is that [now] a highly interconnected world, to a great extent on social media, is setting the stage for distributing information and misinformation about COVID‐19.”

Consider the following composite vignette: Mr. A is a 70-year-old retiree with a history of COPD who has been advised by his PCP to get vaccinated against COVID-19. He is extremely reluctant to do so, fearing that “the vaccine is going to change my DNA” and “might even give me COVID.” He has heard from friends on social media that vaccine developers “faked the results” and are “in cahoots with the federal government.” Mr. A has heard “experts” declare the vaccines safe, but does not trust them. Mr. A has no psychiatric or substance abuse history, and there are no cognitive, perceptual, or other abnormalities in Mr. A’s mental status exam.

Mr. A’s beliefs qualify as a “conspiracy theory,” but probably represent widely held misconceptions about COVID-19 vaccines, as well as widespread mistrust of pharmaceutical companies and the federal government. Based on the information provided, there is no basis for concluding that Mr. A is psychotic or delusional. His beliefs appear to be the result of “epistemic mistrust” of authoritative informational accounts, biased information processing, and exposure to misinformation.

How should the physician manage and care for patients like Mr. A? Absent frank delusions, there is no role for antipsychotic medication, though for extremely anxious patients, a time-limited course of an antianxiety agent may sometimes be warranted. In addition to providing accurate medical information to the patient, the physician should avoid arguing, or trying to “talk the patient out of” his or her belief. Instead, the focus should be on sustaining and strengthening the physician-patient alliance, establishing an atmosphere of respect and safety, clarifying differences in trusted sources of medical information, and allowing the patient time to process the physician’s recommendations.

One-to-one engagement with health care providers has proved effective in reducing vaccine hesitancy and correcting misinformation. For patients with less fixed conspiracy theory beliefs, it may sometimes be helpful to gently offer alternative hypotheses to the patient’s conspiracy theory, using elements of cognitive-behavioral therapy (CBT). For example, a physician might ask, “Is it possible that the online source you read was mistaken about the vaccine changing your DNA?” while reminding patients that – contrary to popular belief – mRNA vaccines have been in development against cancer for several decades.

Challenging beliefs collaboratively and acknowledging areas of uncertainty, rather than confronting or arguing about false beliefs, can foster trust between physician and patient and, at the very least, open a dialogue regarding potential exposure to medical misinformation. “Inoculation” strategies that present and then dispel misinformation before patients become aware of it are among the best supported strategies for mitigating conspiracy theory belief. Ideally, physicians and health care systems should maintain an ongoing “inventory” of medical misinformation circulating online and “beat it to the punch” with reliable information.

Finally, because believing in conspiracy theories is often associated with a sense of uncertainty, and feeling that one’s life is “out of control,” medical interventions can be framed as ways of regaining control and appealing to patients’ values; for example, saying, “By getting the vaccine, you’ll be more likely to stay in good health, protect your family, and do all the things you want to do.”
 

Dr. Pies is professor of psychiatry and a lecturer on bioethics and humanities at State University of New York, Syracuse. Dr. Pierre is a health sciences clinical professor in the department of psychiatry and biobehavioral sciences at the University of California, Los Angeles. A version of this article first appeared on Medscape.com.

When many people across the country, not to mention in public office, believe that the world is run by a group of Satanic pedophiles that includes top Democrats and Hollywood elites, and that former President Trump is leading a secret mission to bring these evildoers to justice, one can’t help but ask if they’re at least to some degree mentally impaired.

Conspiracy theories are often received with psychiatric connotations; associated with paranoid plan-hatchers, and nonbelieving outsiders. But whereas theories such as QAnon strain credibility for many people, we would argue that they are likely not the product of psychosis or mental illness; nor do conspiracy theories in general represent delusions.

For one thing, surveys have consistently revealed that about 50% of the population believes in at least one conspiracy theory. Furthermore, there are several substantive differences between conspiracy theory beliefs and delusions.

Some researchers consider conspiracy theories to be “a subset of false beliefs,” but  most scholars, ourselves included, do not prejudge their validity or veracity. Real-life conspiracies, such as the CIA’s MK-Ultra program, have clearly occurred throughout history.

Our central contention is that belief in conspiracy theories is distinct from psychosis, and more closely resembles extreme but subculturally sanctioned religious or political beliefs. However, the line between believing in conspiracies and being delusional becomes blurred when the believer becomes part of the conspiracy theory and feels compelled to act on the belief as part of a personal mission.

Take Edgar Maddison Welch, a 28-year-old man who firmly believed the so-called “Pizzagate” conspiracy theory – the baseless claim that Hillary Clinton and Democratic elites were running a child sex-trafficking ring out of a Washington, DC, pizzeria. Seeing himself a potential savior of children, Mr. Welch drove 350 miles to the pizza shop from his home in North Carolina in December 2016 and fired three shots from an AR-15 style rifle into a locked closet door, ultimately surrendering to police. However, on questioning he quickly conceded, “The intel on this wasn’t 100%.”

Who believes in conspiracy theories?

Given that half the population believes in at least one conspiracy theory, it should come as little surprise that there is no reliable “profile” for believers. Although some studies have suggested associations with low education, right-wing political orientation, and certain personality traits like subclinical paranoia and schizotypy, such findings have been inconsistent and may vary across specific conspiracy theory. Associations between conspiracy belief and paranoia suggest overlap within a “conspiratorial mindset,” with recent evidence that “distrust of officialdom” is a key mediator between believing in conspiracies and political ideology.

Other quantitative “cognitive quirks” reported in those who believe in conspiracies are a need for certainty and control, a need for uniqueness, illusory pattern perception, and lack of analytical thinking. It’s unclear which of these factors may represent universal cognitive explanations for conspiratorial beliefs, vs. those that might be related to specific beliefs, such as the need for certainty during times of crisis and societal upheaval, when conspiracy theories tend to flourish.

Much of the research on conspiracy theory belief is based on the questionable premise that it’s best understood at the level of the individual’s psychopathology, or the “deficit model,” as it’s called. One of us (JMP) has instead proposed a two-component model that includes social and informational contexts. The first component – epistemic mistrust – involves mistrusting conventional, “authoritative” knowledge. The second involves biased information processing and exposure to misinformation, often transmitted by word of mouth, or through social networks. With this model, believing in conspiracy theories could be conceived as involving “delusion-like beliefs,” but not frank psychosis or full-blown delusions, as one might see, for example, in schizophrenia.

Indeed, many of the cognitive characteristics associated with conspiracy theory belief are universal, continuously distributed traits, varying in quantity, rather than all-or-none variables or distinct symptoms of mental illness.

Essentially, delusions are fixed, false, usually unshared beliefs, often based on subjective “inner” experience. (One rare exception is the so-called folie à deux, in which two people appear to “share” the same delusion; however, psychiatrists have long debated whether both individuals should be considered truly delusional). The delusion’s content is often “self-referential”; i.e., focused primarily on the believer.

In contrast, conspiracy theories are usually, but not necessarily, false. They are typically shared beliefs that don’t explicitly or directly involve the believer, and are based on evidence that one finds “out there,” such as on the Internet. This speaks to the highly communal nature of so many conspiracy theories – networks of like-minded individuals reinforcing one another’s beliefs in a particular socio-cultural context.
 

Conspiracy theory belief, COVID-19, and medical intervention

As for medical conspiracy theories, none have flourished recently more so than those involving the COVID-19 pandemic. As a recent editorial by Stein and colleagues noted, “Some conspiratorial claims include assertions that COVID‐19 is a hoax; arguments that the virus was created artificially and spread on purpose as a bioweapon; or allegations that governments are using the emergency situation to pursue their antidemocratic goals. … Other conspiracies argued that people in power are taking advantage of the pandemic as a plan to inject microchip quantum-dot spy software and monitor people.”

Stein and colleagues make the important point that a “key difference between COVID‐19 and the 1918 flu pandemic ... is that [now] a highly interconnected world, to a great extent on social media, is setting the stage for distributing information and misinformation about COVID‐19.”

Consider the following composite vignette: Mr. A is a 70-year-old retiree with a history of COPD who has been advised by his PCP to get vaccinated against COVID-19. He is extremely reluctant to do so, fearing that “the vaccine is going to change my DNA” and “might even give me COVID.” He has heard from friends on social media that vaccine developers “faked the results” and are “in cahoots with the federal government.” Mr. A has heard “experts” declare the vaccines safe, but does not trust them. Mr. A has no psychiatric or substance abuse history, and there are no cognitive, perceptual, or other abnormalities in Mr. A’s mental status exam.

Mr. A’s beliefs qualify as a “conspiracy theory,” but probably represent widely held misconceptions about COVID-19 vaccines, as well as widespread mistrust of pharmaceutical companies and the federal government. Based on the information provided, there is no basis for concluding that Mr. A is psychotic or delusional. His beliefs appear to be the result of “epistemic mistrust” of authoritative informational accounts, biased information processing, and exposure to misinformation.

How should the physician manage and care for patients like Mr. A? Absent frank delusions, there is no role for antipsychotic medication, though for extremely anxious patients, a time-limited course of an antianxiety agent may sometimes be warranted. In addition to providing accurate medical information to the patient, the physician should avoid arguing, or trying to “talk the patient out of” his or her belief. Instead, the focus should be on sustaining and strengthening the physician-patient alliance, establishing an atmosphere of respect and safety, clarifying differences in trusted sources of medical information, and allowing the patient time to process the physician’s recommendations.

One-to-one engagement with health care providers has proved effective in reducing vaccine hesitancy and correcting misinformation. For patients with less fixed conspiracy theory beliefs, it may sometimes be helpful to gently offer alternative hypotheses to the patient’s conspiracy theory, using elements of cognitive-behavioral therapy (CBT). For example, a physician might ask, “Is it possible that the online source you read was mistaken about the vaccine changing your DNA?” while reminding patients that – contrary to popular belief – mRNA vaccines have been in development against cancer for several decades.

Challenging beliefs collaboratively and acknowledging areas of uncertainty, rather than confronting or arguing about false beliefs, can foster trust between physician and patient and, at the very least, open a dialogue regarding potential exposure to medical misinformation. “Inoculation” strategies that present and then dispel misinformation before patients become aware of it are among the best supported strategies for mitigating conspiracy theory belief. Ideally, physicians and health care systems should maintain an ongoing “inventory” of medical misinformation circulating online and “beat it to the punch” with reliable information.

Finally, because believing in conspiracy theories is often associated with a sense of uncertainty, and feeling that one’s life is “out of control,” medical interventions can be framed as ways of regaining control and appealing to patients’ values; for example, saying, “By getting the vaccine, you’ll be more likely to stay in good health, protect your family, and do all the things you want to do.”
 

Dr. Pies is professor of psychiatry and a lecturer on bioethics and humanities at State University of New York, Syracuse. Dr. Pierre is a health sciences clinical professor in the department of psychiatry and biobehavioral sciences at the University of California, Los Angeles. A version of this article first appeared on Medscape.com.

When many people across the country, not to mention in public office, believe that the world is run by a group of Satanic pedophiles that includes top Democrats and Hollywood elites, and that former President Trump is leading a secret mission to bring these evildoers to justice, one can’t help but ask if they’re at least to some degree mentally impaired.

Conspiracy theories are often received with psychiatric connotations; associated with paranoid plan-hatchers, and nonbelieving outsiders. But whereas theories such as QAnon strain credibility for many people, we would argue that they are likely not the product of psychosis or mental illness; nor do conspiracy theories in general represent delusions.

For one thing, surveys have consistently revealed that about 50% of the population believes in at least one conspiracy theory. Furthermore, there are several substantive differences between conspiracy theory beliefs and delusions.

Some researchers consider conspiracy theories to be “a subset of false beliefs,” but  most scholars, ourselves included, do not prejudge their validity or veracity. Real-life conspiracies, such as the CIA’s MK-Ultra program, have clearly occurred throughout history.

Our central contention is that belief in conspiracy theories is distinct from psychosis, and more closely resembles extreme but subculturally sanctioned religious or political beliefs. However, the line between believing in conspiracies and being delusional becomes blurred when the believer becomes part of the conspiracy theory and feels compelled to act on the belief as part of a personal mission.

Take Edgar Maddison Welch, a 28-year-old man who firmly believed the so-called “Pizzagate” conspiracy theory – the baseless claim that Hillary Clinton and Democratic elites were running a child sex-trafficking ring out of a Washington, DC, pizzeria. Seeing himself a potential savior of children, Mr. Welch drove 350 miles to the pizza shop from his home in North Carolina in December 2016 and fired three shots from an AR-15 style rifle into a locked closet door, ultimately surrendering to police. However, on questioning he quickly conceded, “The intel on this wasn’t 100%.”

Who believes in conspiracy theories?

Given that half the population believes in at least one conspiracy theory, it should come as little surprise that there is no reliable “profile” for believers. Although some studies have suggested associations with low education, right-wing political orientation, and certain personality traits like subclinical paranoia and schizotypy, such findings have been inconsistent and may vary across specific conspiracy theory. Associations between conspiracy belief and paranoia suggest overlap within a “conspiratorial mindset,” with recent evidence that “distrust of officialdom” is a key mediator between believing in conspiracies and political ideology.

Other quantitative “cognitive quirks” reported in those who believe in conspiracies are a need for certainty and control, a need for uniqueness, illusory pattern perception, and lack of analytical thinking. It’s unclear which of these factors may represent universal cognitive explanations for conspiratorial beliefs, vs. those that might be related to specific beliefs, such as the need for certainty during times of crisis and societal upheaval, when conspiracy theories tend to flourish.

Much of the research on conspiracy theory belief is based on the questionable premise that it’s best understood at the level of the individual’s psychopathology, or the “deficit model,” as it’s called. One of us (JMP) has instead proposed a two-component model that includes social and informational contexts. The first component – epistemic mistrust – involves mistrusting conventional, “authoritative” knowledge. The second involves biased information processing and exposure to misinformation, often transmitted by word of mouth, or through social networks. With this model, believing in conspiracy theories could be conceived as involving “delusion-like beliefs,” but not frank psychosis or full-blown delusions, as one might see, for example, in schizophrenia.

Indeed, many of the cognitive characteristics associated with conspiracy theory belief are universal, continuously distributed traits, varying in quantity, rather than all-or-none variables or distinct symptoms of mental illness.

Essentially, delusions are fixed, false, usually unshared beliefs, often based on subjective “inner” experience. (One rare exception is the so-called folie à deux, in which two people appear to “share” the same delusion; however, psychiatrists have long debated whether both individuals should be considered truly delusional). The delusion’s content is often “self-referential”; i.e., focused primarily on the believer.

In contrast, conspiracy theories are usually, but not necessarily, false. They are typically shared beliefs that don’t explicitly or directly involve the believer, and are based on evidence that one finds “out there,” such as on the Internet. This speaks to the highly communal nature of so many conspiracy theories – networks of like-minded individuals reinforcing one another’s beliefs in a particular socio-cultural context.
 

Conspiracy theory belief, COVID-19, and medical intervention

As for medical conspiracy theories, none have flourished recently more so than those involving the COVID-19 pandemic. As a recent editorial by Stein and colleagues noted, “Some conspiratorial claims include assertions that COVID‐19 is a hoax; arguments that the virus was created artificially and spread on purpose as a bioweapon; or allegations that governments are using the emergency situation to pursue their antidemocratic goals. … Other conspiracies argued that people in power are taking advantage of the pandemic as a plan to inject microchip quantum-dot spy software and monitor people.”

Stein and colleagues make the important point that a “key difference between COVID‐19 and the 1918 flu pandemic ... is that [now] a highly interconnected world, to a great extent on social media, is setting the stage for distributing information and misinformation about COVID‐19.”

Consider the following composite vignette: Mr. A is a 70-year-old retiree with a history of COPD who has been advised by his PCP to get vaccinated against COVID-19. He is extremely reluctant to do so, fearing that “the vaccine is going to change my DNA” and “might even give me COVID.” He has heard from friends on social media that vaccine developers “faked the results” and are “in cahoots with the federal government.” Mr. A has heard “experts” declare the vaccines safe, but does not trust them. Mr. A has no psychiatric or substance abuse history, and there are no cognitive, perceptual, or other abnormalities in Mr. A’s mental status exam.

Mr. A’s beliefs qualify as a “conspiracy theory,” but probably represent widely held misconceptions about COVID-19 vaccines, as well as widespread mistrust of pharmaceutical companies and the federal government. Based on the information provided, there is no basis for concluding that Mr. A is psychotic or delusional. His beliefs appear to be the result of “epistemic mistrust” of authoritative informational accounts, biased information processing, and exposure to misinformation.

How should the physician manage and care for patients like Mr. A? Absent frank delusions, there is no role for antipsychotic medication, though for extremely anxious patients, a time-limited course of an antianxiety agent may sometimes be warranted. In addition to providing accurate medical information to the patient, the physician should avoid arguing, or trying to “talk the patient out of” his or her belief. Instead, the focus should be on sustaining and strengthening the physician-patient alliance, establishing an atmosphere of respect and safety, clarifying differences in trusted sources of medical information, and allowing the patient time to process the physician’s recommendations.

One-to-one engagement with health care providers has proved effective in reducing vaccine hesitancy and correcting misinformation. For patients with less fixed conspiracy theory beliefs, it may sometimes be helpful to gently offer alternative hypotheses to the patient’s conspiracy theory, using elements of cognitive-behavioral therapy (CBT). For example, a physician might ask, “Is it possible that the online source you read was mistaken about the vaccine changing your DNA?” while reminding patients that – contrary to popular belief – mRNA vaccines have been in development against cancer for several decades.

Challenging beliefs collaboratively and acknowledging areas of uncertainty, rather than confronting or arguing about false beliefs, can foster trust between physician and patient and, at the very least, open a dialogue regarding potential exposure to medical misinformation. “Inoculation” strategies that present and then dispel misinformation before patients become aware of it are among the best supported strategies for mitigating conspiracy theory belief. Ideally, physicians and health care systems should maintain an ongoing “inventory” of medical misinformation circulating online and “beat it to the punch” with reliable information.

Finally, because believing in conspiracy theories is often associated with a sense of uncertainty, and feeling that one’s life is “out of control,” medical interventions can be framed as ways of regaining control and appealing to patients’ values; for example, saying, “By getting the vaccine, you’ll be more likely to stay in good health, protect your family, and do all the things you want to do.”
 

Dr. Pies is professor of psychiatry and a lecturer on bioethics and humanities at State University of New York, Syracuse. Dr. Pierre is a health sciences clinical professor in the department of psychiatry and biobehavioral sciences at the University of California, Los Angeles. A version of this article first appeared on Medscape.com.

For more than half a century, haloperidol has been used as a first-line medication for psychiatric agitation constituting a “behavioral emergency” when a patient cannot or will not take oral medication. Today, haloperidol is most commonly administered as an IM injection along with an anticholinergic medication to minimize extrapyramidal symptoms (EPS) and a benzodiazepine for additional sedation. The multiple-medication “cocktail” is often referred to by double-entendre nicknames, such as “B-52” or “5250” (ie, haloperidol, 5 mg; lorazepam, 2 mg; and diphenhydramine, 50 mg). In this article, I discuss whether haloperidol, a first-generation antipsychotic (FGA) medication developed in 1958, still deserves to be the IM “gold standard” for managing emergency psychiatric agitation.

Earlier evidence of haloperidol’s efficacy

The initial “discovery” of antipsychotic medications was made in 1951 based on the inadvertent observation that chlorpromazine had the potential to calm surgical patients with autonomic activation. This calming effect, described as “désintéressment” (meaning a kind of “indifference to the world”),¹ resulted in a new class of medications replacing barbiturates and bromides as go-to options to achieve “rapid tranquilization” of psychiatric agitation.² Although the ability of antipsychotic medications to gradually reduce positive symptoms, such as delusions and hallucinations, has been attributed to dopamine (D2) antagonism, their more immediate sedating and anti-agitation effects are the result of broader effects as histamine (H1) and alpha-1 adrenergic antagonists.

In the 1970s, haloperidol emerged as a first-line option to manage agitation due to its IM and IV availability, as well as its relative lack of sedation and orthostasis compared with low-potency D2 antagonists such as chlorpromazine. However, haloperidol was observed to have a significant risk of acute EPS, including dystonic reactions.² From the 1970s to the 1990s, numerous prospective clinical trials of haloperidol for the treatment of acute psychotic agitation, including several randomized controlled trials (RCTs) comparing haloperidol to lorazepam, were conducted.³ The design and outcomes of the haloperidol vs lorazepam RCTs were fairly consistent^4-7:

• adult participants with acute agitation and a variety of psychiatric diagnoses, for whom informed consent often was waived due to agitation severity
• randomization to either IM haloperidol, 5 mg, or IM lorazepam, 2 mg, administered every 30 minutes until agitation resolved
• behavioral outcomes measured over several hours using various rating scales, without consistent assessment of EPS
• equivalent efficacy of haloperidol and lorazepam, with symptom resolution usually achieved after 1 to 2 doses (in 30 to 60 minutes), but sometimes longer
• anticholinergic “rescue” allowed for EPS, but not administered prophylactically
• EPS, including dystonia and akathisia, were significantly more frequent with haloperidol compared with lorazepam.⁸

In recognition of the greater risk of EPS with haloperidol compared with lorazepam, and the fact that most study participants were already taking standing doses of antipsychotic medications, some researchers have recommended using benzodiazepines alone as the optimal treatment for agitation.^4,9 A 2012 Cochrane review concluded that the involuntary use of haloperidol alone “could be considered unethical.”^10,11 However, other studies that examined the combination of haloperidol and lorazepam compared with either medication alone found that the combination of the 2 medications was associated with a more rapid resolution of symptoms, which suggests a superior synergistic effect.^6,7,12 By the late 1990s, combined haloperidol and lorazepam, often mixed within a single injection, became the most common strategy to achieve rapid tranquilization in the psychiatric emergency setting.¹³ However, while the combination has been justified as a way to reduce the antipsychotic medication dose and EPS risk,² few studies have compared combinations containing <5 mg of haloperidol. As a result, the apparent superiority of combined haloperidol and lorazepam compared with either medication alone may be a simple cumulative dose effect rather than true synergism. It is also important to note that adding lorazepam to haloperidol does not mitigate the risk of EPS such as dystonia in the absence of anticholinergic medication.⁸ To date, however, there have been no clinical trials investigating the efficacy of IM haloperidol, lorazepam, and benztropine or diphenhydramine given together.

Newer RCTs tell a different story

With the availability of second-generation antipsychotics (SGAs) in IM formulations, clinical trials over the past 2 decades have focused on comparing SGAs with haloperidol alone as the “gold standard” control for acute agitation. Compared with previous trials of haloperidol vs lorazepam, these clinical trials of SGAs vs haloperidol included^8,14-22:

• Study participants who signed informed consent (and were likely less agitated)
• IM haloperidol doses typically >5 mg (eg, 6.5 to 10 mg).

As with studies comparing lorazepam with haloperidol, the results of these RCTs revealed that IM aripiprazole, olanzapine, and ziprasidone were at least as effective as IM haloperidol, with haloperidol having a significantly increased risk of akathisia, dystonia, and other EPS.^8,14-22 The greater EPS risk of haloperidol is not surprising given the use of comparison doses up to 10 mg.

An updated 2017 Cochrane review of haloperidol for psychosis-induced aggression or agitation concluded that⁹:

• haloperidol is an effective intervention, although the evidence is “weak”
• significant treatment effects may take as long as 1 to 2 hours following multiple IM injections
• in contrast to SGAs, treatment with haloperidol carries a significant risk of EPS
• adding a benzodiazepine “does not have strong evidence of benefit and carries risk of additional harm.”

Continue to: Haloperidol's well-known toxicity

Haloperidol has been associated with numerous adverse effects:

Akathisia and other acute EPS. Treatment with even a single dose of IM haloperidol can result in acute EPS, including dystonia and akathisia. At best, such adverse effects are subjectively troubling and unpleasant; at worst, akathisia can exacerbate and be mistaken for agitation, leading to administration of more medication²³ and the possible development of suicidal or violent behavior.^24-25 In the studies reviewed above, the overall rate of EPS was as high as 21% after treatment with haloperidol,¹⁶ with parkinsonism occurring in up to 17% of patients,¹⁹ dystonia in up to 11%,⁷ and akathisia in up to 10%.¹⁵ However, because specific EPS were assessed inconsistently, and sometimes not at all, the rate of akathisia—arguably the most relevant and counter-therapeutic adverse effect related to agitation—remains unclear.

In another study that specifically assessed for akathisia in patients treated with haloperidol, up to 40% experienced akathisia 6 hours after a single oral dose of 5 mg.²⁶ Even a single dose of IV prochlorperazine, another dopamine-antagonist routinely used to treat nausea in the emergency department (ED), has been reported to cause akathisia in up to 44% of patients.²⁷ Such results suggest that when akathisia is carefully assessed, the rate with even brief FGA exposure may approach nearly half of treated patients. Because akathisia is typically dose-related, and considering that many patients receiving IM haloperidol may receive multiple injections in addition to standing doses of oral medications, akathisia may be underrecognized in patients who are agitated, with a much greater risk than is generally presumed.

Although anticholinergic medications or benzodiazepinesare often administered as part of a haloperidol “cocktail,” these medications often do not adequately resolve emergent akathisia.^26,28 No clinical trials of IM haloperidol combined with benztropine or diphenhydramine have been published, but several studies suggest that combining haloperidol with promethazine—a phenothiazine with strong antihistaminergic and anticholinergic activity, but only weak antidopaminergic activity—can decrease the risk of dystonia relative to haloperidol alone.^8,22,29,30 However, there have also been reports of promethazine causing dystonia.^31,32 In addition, 1 trial of IM haloperidol, 2.5 mg, combined with promethazine reported that 74% of patients still had at least 1 form of EPS.³⁰ Because the clinical trials of haloperidol with promethazine did not specifically assess for akathisia, promethazine’s ability to decrease the risk of akathisia remains unknown.

Cardiotoxicity. Although low-potency antipsychotic medications such as chlorpromazine are more sedating than haloperidol, the latter is preferred as an IM antipsychotic medication for agitation because of its lower risk of hypotension.² In terms of cardiac effects, all antipsychotic medications carry a risk of QTc prolongation, with possible progression to the potentially lethal arrhythmia torsades de pointes as a result of interference with cardiac potassium channels.³³ In 2007, the FDA added a “black-box” warning about this risk for haloperidol, in the wake of a disproportionately high number of reported cases associated with IV administration, sometimes even after a single dose.³⁴

Continue to: Although there is no direct evidence...

Although there is no direct evidence that the cardiac risks associated with IV haloperidol apply to IM administration, epidemiologic studies indicate that oral haloperidol carries an elevated risk of ventricular arrhythmia and sudden cardiac death,^35,36 with 1 study reporting greater risk compared with other SGAs.³⁷ Haloperidol, whether administered orally or IM, may therefore be an especially poor choice for patients with agitation who are at risk for arrhythmia, including those with relevant medical comorbidities or delirium.³⁴

Neuronal cell death. Several lines of research evidence have demonstrated that haloperidol can cause cellular injury or death in neuronal tissue in a dose-dependent fashion through a variety of mechanisms.³⁸ By contrast, SGAs have been shown to have neuroprotective effects.³⁹ While these findings have mostly come from studies conducted in animals or in vitro human tumor cell lines, some researchers have nonetheless called for haloperidol to be banned, noting that if its neurotoxic effects were more widely known, “we would realize what a travesty it is to use [such] a brain-unfriendly drug.”⁴⁰

Several reasonable alternatives

Echoing the earlier Cochrane review of haloperidol for psychosis-induced aggression or agitation,¹⁰ a 2017 update concluded, “If no other alternative exists, sole use of intramuscular haloperidol could be life-saving. Where additional drugs are available, sole use of haloperidol for extreme emergency could be considered unethical.”⁹

What then are reasonable alternatives to replace IM haloperidol for agitation? Clinicians should consider the following nonpharmacologic and pharmacologic interventions:

Nonpharmacologic interventions. Several behavioral interventions have been demonstrated to be effective for managing acute agitation, including verbal de-escalation, enhanced “programming” on the inpatient units, and the judicious use of seclusion.^41-43 While such interventions may demand additional staff or resources, they have the potential to lower long-term costs, reduce injuries to patients and staff, and improve the quality of care.⁴³ The use of IM haloperidol as a form of “chemical restraint” does not represent standard-of-care treatment,³ and from an ethical perspective, should never be implemented punitively or to compensate for substandard care in the form of inadequate staffing or staff training.

Continue to: Benzodiazepines

Benzodiazepines. Lorazepam offers an attractive alternative to haloperidol without the risk of EPS.^2,4,8 However, lorazepam alone may be perceived as less efficacious than a haloperidol “cocktail” because it represents less overall medication. Some evidence has suggested that lorazepam, 4 mg, might be the most appropriate dose, although it has only rarely been studied in clinical trials of acute agitation.³ Midazolam is another IM benzodiazepine alternative to IM haloperidol that has been shown to achieve more rapid sedation than either haloperidol or lorazepam,^44,45 although it can cause substantial anterograde amnesia and also has an FDA black-box warning for respiratory depression associated with IV administration.

Respiratory depression is frequently cited as an argument against using lorazepam for agitation, as if the therapeutic window is extremely narrow with ineffectiveness at 2 mg, but potential lethality beyond that dose. In fact, serious respiratory depression with lorazepam is unlikely in the absence of chronic obstructive pulmonary disease (COPD), obstructive sleep apnea, or concomitant alcohol or other sedative use.⁴⁶ Case reports have documented therapeutic lorazepam dosing of 2 to 4 mg every 2 hours up to 20 to 30 mg/d in patients with manic agitation.⁴⁷ Even in patients with COPD, significant respiratory depression tends not to occur at doses <8 mg.⁴⁸ A more evidence-based concern about lorazepam dosing is that 2 mg might be ineffective in patients with established tolerance. For example, 1 report described a patient in acute alcohol withdrawal who required dosing lorazepam to 1,600 mg within 24 hours.⁴⁹ Collectively, these reports suggest that lorazepam has a much wider therapeutic window than is typically perceived, and that dosing with 3 to 4 mg IM is a reasonable option for agitation when 2 mg is likely to be inadequate.

Paradoxical disinhibition is another concern that might prevent benzodiazepines from being used alone as a first-line intervention for emergency treatment of agitation. However, similar to respiratory depression, this adverse event is relatively rare and tends to occur in children and geriatric patients, individuals intoxicated with alcohol or other sedatives, and patients with brain injury, developmental delay, or dementia.^23,46 Although exacerbation of aggression has not been demonstrated in the RCTs examining benzodiazepines for agitation reviewed above, based on other research, some clinicians have expressed concerns about the potential for benzodiazepines to exacerbate aggression in patients with impulse control disorders and a history of violent behavior.⁵⁰

The 2005 Expert Consensus Panel for Behavioral Emergencies⁵¹ recommended the use of lorazepam alone over haloperidol for agitation for patients for whom the diagnosis is unknown or includes the following:

• stimulant intoxication
• personality disorder
• comorbid obesity
• comorbid cardiac arrhythmia
• a history of akathisia and other EPS
• a history of amenorrhea/galactorrhea
• a history of seizures.

In surveys, patients have ranked lorazepam as the preferred medication for emergency agitation, whereas haloperidol was ranked as one of the least-preferred options.^51,52

Continue to: Second-generation antipsychotics

Second-generation antipsychotics. The SGAs available in IM formulations, such as aripiprazole, olanzapine, and ziprasidone, have been shown to be at least as effective as haloperidol for the treatment of acute agitation (in 2015, the short-acting injectable formulation of aripiprazole was discontinued in the United States independent of safety or efficacy issues⁵³). A review of RCTs examining IM SGAs for the treatment of agitation concluded that the number needed to treat for response compared with placebo was 5 for aripiprazole, 3 for olanzapine, and 3 for ziprasidone.⁵⁴ In terms of safety, a meta-analysis of studies examining IM medications for agitation confirmed that the risk of acute EPS, including dystonia, akathisia, and parkinsonism, is significantly lower with SGAs compared with haloperidol.⁵⁵ An RCT comparing IM ziprasidone with haloperidol found equivalently modest effects on QTc prolongation.⁵⁶ Therefore, SGAs are an obvious and evidence-based option for replacing haloperidol as a treatment for acute agitation.

Unfortunately, for clinicians hoping to replace haloperidol within a multiple-medication IM “cocktail,” there have been no published controlled trials of SGAs combined with benzodiazepines. Although a short report indicated that aripiprazole and lorazepam are chemically compatible to be combined within a single injection,⁵⁷ the package insert for aripiprazole warns that “If parenteral benzodiazepine therapy is deemed necessary in addition to ABILIFY injection treatment, patients should be monitored for excessive sedation and for orthostatic hypotension.”⁵⁸ The package insert for olanzapine likewise lists the combination of lorazepam and olanzapine as a drug interaction that can potentiate sedation, and the manufacturer issued specific warnings about parenteral combination.^59,60 A single published case of significant hypotension with combined IM olanzapine and lorazepam,⁶⁰ together with the fact that IM olanzapine can cause hypotension by itself,⁶¹ has discouraged the coadministration of these medications. Nonetheless, the combination is used in some emergency settings, with several retro­spective studies failing to provide evidence of hypotension or respiratory depression as adverse effects.^62-64

Droperidol. Droperidol was formerly a popular choice for managing acute agitation, with evidence from RCTs that droperidol, 5 mg, can improve symptoms significantly faster than either haloperidol, 5 mg, or lorazepam, 2 mg, and is absorbed just as rapidly whether administered IV or IM.^65-67 However, a 2001 FDA black-box warning about QTc prolongation included recommendations that a screening electrocardiogram should be obtained before administering droperidol. This action greatly curtailed the use of droperidol, and for some time, it was not marketed or available in the United States.

Over the past decade, however, droperidol has returned to the US market⁶⁸ and its IV and IM usage has been revitalized for managing patients with agitation within or en route to the ED. Studies have demonstrated droperidol efficacy comparable to midazolam, ziprasidone, or olanzapine, as well as effectiveness as an IV adjunct to midazolam.^69-71 In contrast to the FDA black-box warning, retrospective studies and RCTs of both IV and IM droperidol suggest that QTc prolongation and torsades de pointes are rare events that do not occur any more frequently than they do with haloperidol, even at doses >10 mg.^72,73 However, in studies involving patients with drug intoxication and treatment with multiple medications, oversedation to the point of needing rescue intervention was reported. In an emergency setting where these issues are relatively easily managed, such risks may be better tolerated than in psychiatric settings.

With earlier studies examining the use of droperidol in an acute psychiatric setting that reported a more rapid onset of action than haloperidol,^65-67 a 2016 Cochrane review concluded that there was high-quality evidence to support droperidol’s use for psychosis-induced agitation.⁷⁴ However, a 2015 RCT comparing IM droperidol, 10 mg, to haloperidol, 10 mg, found equivalent efficacy and response times (with maximal response occurring within 2 hours) and concluded that droperidol had no advantage over haloperidol.⁷⁵ Because none of the clinical trials that evaluated droperidol have included assessments for EPS, its risk of akathisia remains uncertain.

Continue to: Ketamine

Ketamine. In recent years, ketamine has been used to treat acute agitation within or en route to the ED. Preliminary observational studies support ketamine’s efficacy when administered via IV or IM routes,⁷⁶ with more rapid symptomatic improvement compared with haloperidol, lorazepam, or midazolam alone.⁷⁷ Reported adverse effects of ketamine include dissociation, psychotic exacerbation, and respiratory depression,⁷⁶ although 1 small naturalistic study found no evidence of exacerbation of psychotic or other psychiatric symptoms.⁷⁸ An ongoing RCT is comparing IM ketamine, 5 mg/kg, to combined IM haloperidol, 5 mg, and midazolam, 5 mg.⁷⁹ Although various ketamine formulations are increasingly being used in psychiatry, active psychosis is generally regarded as a contraindication. It is premature to recommend parenteral ketamine administration for agitation within most psychiatric settings until more research on safety has been completed.

Haloperidol, or something else? Practical considerations

Consider the following factors when deciding whether to use haloperidol or one of its alternatives:

Limitations of the evidence. Modern clinical trials requiring informed consent often do not include the kind of severe agitation that clinicians encounter in acute psychiatric, emergency, or forensic settings. In addition, standard interventions, such as 3-medication haloperidol “cocktails,” have not been evaluated in clinical trials. Clinicians are therefore often in the dark about optimal evidence-based practices.

Treatment goals. Psychiatric agitation has many causes, with a range of severity that warrants a commensurate range of responses. Protocols for managing acute agitation should include graded interventions that begin with nonpharmacologic interventions and voluntary oral medications, and move to involuntary IM medications when necessary.

While treatment guidelines clearly recommend against IM medications as “chemical restraint” with a goal of sedating a patient until he/she is unconscious,^3,51 such outcomes are nonetheless often sought by staff who are concerned about the risk of injuries during a behavioral emergency. In such instances, the risks of violence towards patients and staff may outweigh concerns about adverse effects in a risk-benefit analysis. Consequently, clinicians may be prone to “skip over” graded interventions because they assume they “won’t work” in favor of administering involuntary multiple-medication haloperidol “cocktails” despite risks of excess sedation, EPS, and cardiotoxicity. Treatment settings should critically evaluate such biased preferences, with a goal of developing tailored, evidence-based strategies that maximize benefits while minimizing excess sedation and other untoward adverse effects, with an eye towards promoting better overall patient care and reducing length of stay.^42,43,80

Continue to: Limitations of available medications

Limitations of available medications. There is no perfect medication for the management of acute agitation. Evidence indicates that pharmacologic options take 15 minutes to several hours to resolve acute agitation, even potentially more rapid-acting medications such as midazolam and droperidol. This is well beyond most clinicians’ desired window for response time in a behavioral emergency. Multiple-medication “cocktails” may be used with the hope of hastening response time, but may not achieve this goal at the expense of increasing the risk of adverse effects and the likelihood that a patient will remain sedated for a prolonged time. In the real world, this often means that by the time a psychiatrist comes to evaluate a patient who has been given emergency medications, the patient cannot be aroused for an interview. Ideally, medications would calm an agitated patient rapidly, without excess or prolonged sedation.⁸⁰ Less-sedating SGAs, such as ziprasidone, might have this potential, but can sometimes be perceived as ineffective.

Avoiding akathisia. Akathisia’s potential to worsen and be mistaken for agitation makes it an especially concerning, if underappreciated, adverse effect of haloperidol that is often not adequately assessed in clinical trials or practice. In light of evidence that akathisia can occur in nearly half of patients receiving a single 5 mg-dose of haloperidol, it is difficult to justify the use of this medication for agitation when equally effective options exist with a lower risk of EPS.

While haloperidol-induced akathisia could in theory be mitigated by adding anticholinergic medications or benzodiazepines, previous studies have found that such strategies have limited effectiveness compared to “gold standard” treatment with propranolol.^28,81,82 Furthermore, the half-lives of anticholinergic medications, such as benztropine or diphenhydramine, are significantly shorter than that of a single dose of haloperidol, which can be as long as 37 hours.⁸³ Therefore, akathisia and other EPS could emerge or worsen several hours or even days after receiving an IM haloperidol “cocktail” as the shorter-acting medications wear off. Akathisia is best minimized by avoiding FGAs, such as haloperidol, when treating acute agitation.

Promoting adherence. Although haloperidol is often recommended for acute agitation in patients with schizophrenia or bipolar disorder on the basis that it would treat the underlying condition, many patients who receive IM medications for acute agitation are already prescribed standing doses of oral medication, which increases the risk of cumulative toxicity. In addition, receiving a medication likely to cause acute EPS that is ranked near the bottom of patient preferences may erode the potential for a therapeutic alliance and hamper longer-term antipsychotic medication adherence.

Time for a change

For nearly half a century, haloperidol has been a “gold standard” intervention for IM control in patients with agitation. However, given its potential to produce adverse effects, including a significant risk of akathisia that can worsen agitation, along with the availability of newer pharmacologic options that are at least as effective (Table 1, and Table 2), haloperidol should be retired as a first-line medication for the treatment of agitation. Clinicians would benefit from RCTs investigating the safety and efficacy of novel interventions including frequently-used, but untested medication combinations, as well as nonpharmacologic interventions.

Continue to: Bottom Line

Although there is no perfect IM medication to treat acute agitation, haloperidol’s higher risk of adverse effects relative to newer alternatives suggest that it should no longer be considered a first-line intervention.

Related Resources

• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 1: onset of efficacy. J Emerg Med. 2018;54(3):364-374.
• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 2: safety. J Emerg Med. 2018;54(4): 522-532.

Drug Brand Names

Aripiprazole • Abilify
Benztropine • Cogentin
Chlorpromazine • Thorazine
Diphenhydramine • Benadryl
Droperidol • Inapsine
Haloperidol • Haldol
Ketamine • Ketalar
Lorazepam • Ativan
Midazolam • Versed
Olanzapine • Zyprexa
Prochlorperazine • Compazine
Promethazine • Phenergan
Propranolol • Inderal, Pronol
Ziprasidone • Geodon

For more than half a century, haloperidol has been used as a first-line medication for psychiatric agitation constituting a “behavioral emergency” when a patient cannot or will not take oral medication. Today, haloperidol is most commonly administered as an IM injection along with an anticholinergic medication to minimize extrapyramidal symptoms (EPS) and a benzodiazepine for additional sedation. The multiple-medication “cocktail” is often referred to by double-entendre nicknames, such as “B-52” or “5250” (ie, haloperidol, 5 mg; lorazepam, 2 mg; and diphenhydramine, 50 mg). In this article, I discuss whether haloperidol, a first-generation antipsychotic (FGA) medication developed in 1958, still deserves to be the IM “gold standard” for managing emergency psychiatric agitation.

Earlier evidence of haloperidol’s efficacy

The initial “discovery” of antipsychotic medications was made in 1951 based on the inadvertent observation that chlorpromazine had the potential to calm surgical patients with autonomic activation. This calming effect, described as “désintéressment” (meaning a kind of “indifference to the world”),¹ resulted in a new class of medications replacing barbiturates and bromides as go-to options to achieve “rapid tranquilization” of psychiatric agitation.² Although the ability of antipsychotic medications to gradually reduce positive symptoms, such as delusions and hallucinations, has been attributed to dopamine (D2) antagonism, their more immediate sedating and anti-agitation effects are the result of broader effects as histamine (H1) and alpha-1 adrenergic antagonists.

In the 1970s, haloperidol emerged as a first-line option to manage agitation due to its IM and IV availability, as well as its relative lack of sedation and orthostasis compared with low-potency D2 antagonists such as chlorpromazine. However, haloperidol was observed to have a significant risk of acute EPS, including dystonic reactions.² From the 1970s to the 1990s, numerous prospective clinical trials of haloperidol for the treatment of acute psychotic agitation, including several randomized controlled trials (RCTs) comparing haloperidol to lorazepam, were conducted.³ The design and outcomes of the haloperidol vs lorazepam RCTs were fairly consistent^4-7:

• adult participants with acute agitation and a variety of psychiatric diagnoses, for whom informed consent often was waived due to agitation severity
• randomization to either IM haloperidol, 5 mg, or IM lorazepam, 2 mg, administered every 30 minutes until agitation resolved
• behavioral outcomes measured over several hours using various rating scales, without consistent assessment of EPS
• equivalent efficacy of haloperidol and lorazepam, with symptom resolution usually achieved after 1 to 2 doses (in 30 to 60 minutes), but sometimes longer
• anticholinergic “rescue” allowed for EPS, but not administered prophylactically
• EPS, including dystonia and akathisia, were significantly more frequent with haloperidol compared with lorazepam.⁸

In recognition of the greater risk of EPS with haloperidol compared with lorazepam, and the fact that most study participants were already taking standing doses of antipsychotic medications, some researchers have recommended using benzodiazepines alone as the optimal treatment for agitation.^4,9 A 2012 Cochrane review concluded that the involuntary use of haloperidol alone “could be considered unethical.”^10,11 However, other studies that examined the combination of haloperidol and lorazepam compared with either medication alone found that the combination of the 2 medications was associated with a more rapid resolution of symptoms, which suggests a superior synergistic effect.^6,7,12 By the late 1990s, combined haloperidol and lorazepam, often mixed within a single injection, became the most common strategy to achieve rapid tranquilization in the psychiatric emergency setting.¹³ However, while the combination has been justified as a way to reduce the antipsychotic medication dose and EPS risk,² few studies have compared combinations containing <5 mg of haloperidol. As a result, the apparent superiority of combined haloperidol and lorazepam compared with either medication alone may be a simple cumulative dose effect rather than true synergism. It is also important to note that adding lorazepam to haloperidol does not mitigate the risk of EPS such as dystonia in the absence of anticholinergic medication.⁸ To date, however, there have been no clinical trials investigating the efficacy of IM haloperidol, lorazepam, and benztropine or diphenhydramine given together.

Newer RCTs tell a different story

With the availability of second-generation antipsychotics (SGAs) in IM formulations, clinical trials over the past 2 decades have focused on comparing SGAs with haloperidol alone as the “gold standard” control for acute agitation. Compared with previous trials of haloperidol vs lorazepam, these clinical trials of SGAs vs haloperidol included^8,14-22:

• Study participants who signed informed consent (and were likely less agitated)
• IM haloperidol doses typically >5 mg (eg, 6.5 to 10 mg).

As with studies comparing lorazepam with haloperidol, the results of these RCTs revealed that IM aripiprazole, olanzapine, and ziprasidone were at least as effective as IM haloperidol, with haloperidol having a significantly increased risk of akathisia, dystonia, and other EPS.^8,14-22 The greater EPS risk of haloperidol is not surprising given the use of comparison doses up to 10 mg.

An updated 2017 Cochrane review of haloperidol for psychosis-induced aggression or agitation concluded that⁹:

• haloperidol is an effective intervention, although the evidence is “weak”
• significant treatment effects may take as long as 1 to 2 hours following multiple IM injections
• in contrast to SGAs, treatment with haloperidol carries a significant risk of EPS
• adding a benzodiazepine “does not have strong evidence of benefit and carries risk of additional harm.”

Continue to: Haloperidol's well-known toxicity

Haloperidol has been associated with numerous adverse effects:

Akathisia and other acute EPS. Treatment with even a single dose of IM haloperidol can result in acute EPS, including dystonia and akathisia. At best, such adverse effects are subjectively troubling and unpleasant; at worst, akathisia can exacerbate and be mistaken for agitation, leading to administration of more medication²³ and the possible development of suicidal or violent behavior.^24-25 In the studies reviewed above, the overall rate of EPS was as high as 21% after treatment with haloperidol,¹⁶ with parkinsonism occurring in up to 17% of patients,¹⁹ dystonia in up to 11%,⁷ and akathisia in up to 10%.¹⁵ However, because specific EPS were assessed inconsistently, and sometimes not at all, the rate of akathisia—arguably the most relevant and counter-therapeutic adverse effect related to agitation—remains unclear.

In another study that specifically assessed for akathisia in patients treated with haloperidol, up to 40% experienced akathisia 6 hours after a single oral dose of 5 mg.²⁶ Even a single dose of IV prochlorperazine, another dopamine-antagonist routinely used to treat nausea in the emergency department (ED), has been reported to cause akathisia in up to 44% of patients.²⁷ Such results suggest that when akathisia is carefully assessed, the rate with even brief FGA exposure may approach nearly half of treated patients. Because akathisia is typically dose-related, and considering that many patients receiving IM haloperidol may receive multiple injections in addition to standing doses of oral medications, akathisia may be underrecognized in patients who are agitated, with a much greater risk than is generally presumed.

Although anticholinergic medications or benzodiazepinesare often administered as part of a haloperidol “cocktail,” these medications often do not adequately resolve emergent akathisia.^26,28 No clinical trials of IM haloperidol combined with benztropine or diphenhydramine have been published, but several studies suggest that combining haloperidol with promethazine—a phenothiazine with strong antihistaminergic and anticholinergic activity, but only weak antidopaminergic activity—can decrease the risk of dystonia relative to haloperidol alone.^8,22,29,30 However, there have also been reports of promethazine causing dystonia.^31,32 In addition, 1 trial of IM haloperidol, 2.5 mg, combined with promethazine reported that 74% of patients still had at least 1 form of EPS.³⁰ Because the clinical trials of haloperidol with promethazine did not specifically assess for akathisia, promethazine’s ability to decrease the risk of akathisia remains unknown.

Cardiotoxicity. Although low-potency antipsychotic medications such as chlorpromazine are more sedating than haloperidol, the latter is preferred as an IM antipsychotic medication for agitation because of its lower risk of hypotension.² In terms of cardiac effects, all antipsychotic medications carry a risk of QTc prolongation, with possible progression to the potentially lethal arrhythmia torsades de pointes as a result of interference with cardiac potassium channels.³³ In 2007, the FDA added a “black-box” warning about this risk for haloperidol, in the wake of a disproportionately high number of reported cases associated with IV administration, sometimes even after a single dose.³⁴

Continue to: Although there is no direct evidence...

Although there is no direct evidence that the cardiac risks associated with IV haloperidol apply to IM administration, epidemiologic studies indicate that oral haloperidol carries an elevated risk of ventricular arrhythmia and sudden cardiac death,^35,36 with 1 study reporting greater risk compared with other SGAs.³⁷ Haloperidol, whether administered orally or IM, may therefore be an especially poor choice for patients with agitation who are at risk for arrhythmia, including those with relevant medical comorbidities or delirium.³⁴

Neuronal cell death. Several lines of research evidence have demonstrated that haloperidol can cause cellular injury or death in neuronal tissue in a dose-dependent fashion through a variety of mechanisms.³⁸ By contrast, SGAs have been shown to have neuroprotective effects.³⁹ While these findings have mostly come from studies conducted in animals or in vitro human tumor cell lines, some researchers have nonetheless called for haloperidol to be banned, noting that if its neurotoxic effects were more widely known, “we would realize what a travesty it is to use [such] a brain-unfriendly drug.”⁴⁰

Several reasonable alternatives

Echoing the earlier Cochrane review of haloperidol for psychosis-induced aggression or agitation,¹⁰ a 2017 update concluded, “If no other alternative exists, sole use of intramuscular haloperidol could be life-saving. Where additional drugs are available, sole use of haloperidol for extreme emergency could be considered unethical.”⁹

What then are reasonable alternatives to replace IM haloperidol for agitation? Clinicians should consider the following nonpharmacologic and pharmacologic interventions:

Nonpharmacologic interventions. Several behavioral interventions have been demonstrated to be effective for managing acute agitation, including verbal de-escalation, enhanced “programming” on the inpatient units, and the judicious use of seclusion.^41-43 While such interventions may demand additional staff or resources, they have the potential to lower long-term costs, reduce injuries to patients and staff, and improve the quality of care.⁴³ The use of IM haloperidol as a form of “chemical restraint” does not represent standard-of-care treatment,³ and from an ethical perspective, should never be implemented punitively or to compensate for substandard care in the form of inadequate staffing or staff training.

Continue to: Benzodiazepines

Benzodiazepines. Lorazepam offers an attractive alternative to haloperidol without the risk of EPS.^2,4,8 However, lorazepam alone may be perceived as less efficacious than a haloperidol “cocktail” because it represents less overall medication. Some evidence has suggested that lorazepam, 4 mg, might be the most appropriate dose, although it has only rarely been studied in clinical trials of acute agitation.³ Midazolam is another IM benzodiazepine alternative to IM haloperidol that has been shown to achieve more rapid sedation than either haloperidol or lorazepam,^44,45 although it can cause substantial anterograde amnesia and also has an FDA black-box warning for respiratory depression associated with IV administration.

Respiratory depression is frequently cited as an argument against using lorazepam for agitation, as if the therapeutic window is extremely narrow with ineffectiveness at 2 mg, but potential lethality beyond that dose. In fact, serious respiratory depression with lorazepam is unlikely in the absence of chronic obstructive pulmonary disease (COPD), obstructive sleep apnea, or concomitant alcohol or other sedative use.⁴⁶ Case reports have documented therapeutic lorazepam dosing of 2 to 4 mg every 2 hours up to 20 to 30 mg/d in patients with manic agitation.⁴⁷ Even in patients with COPD, significant respiratory depression tends not to occur at doses <8 mg.⁴⁸ A more evidence-based concern about lorazepam dosing is that 2 mg might be ineffective in patients with established tolerance. For example, 1 report described a patient in acute alcohol withdrawal who required dosing lorazepam to 1,600 mg within 24 hours.⁴⁹ Collectively, these reports suggest that lorazepam has a much wider therapeutic window than is typically perceived, and that dosing with 3 to 4 mg IM is a reasonable option for agitation when 2 mg is likely to be inadequate.

Paradoxical disinhibition is another concern that might prevent benzodiazepines from being used alone as a first-line intervention for emergency treatment of agitation. However, similar to respiratory depression, this adverse event is relatively rare and tends to occur in children and geriatric patients, individuals intoxicated with alcohol or other sedatives, and patients with brain injury, developmental delay, or dementia.^23,46 Although exacerbation of aggression has not been demonstrated in the RCTs examining benzodiazepines for agitation reviewed above, based on other research, some clinicians have expressed concerns about the potential for benzodiazepines to exacerbate aggression in patients with impulse control disorders and a history of violent behavior.⁵⁰

The 2005 Expert Consensus Panel for Behavioral Emergencies⁵¹ recommended the use of lorazepam alone over haloperidol for agitation for patients for whom the diagnosis is unknown or includes the following:

• stimulant intoxication
• personality disorder
• comorbid obesity
• comorbid cardiac arrhythmia
• a history of akathisia and other EPS
• a history of amenorrhea/galactorrhea
• a history of seizures.

In surveys, patients have ranked lorazepam as the preferred medication for emergency agitation, whereas haloperidol was ranked as one of the least-preferred options.^51,52

Continue to: Second-generation antipsychotics

Second-generation antipsychotics. The SGAs available in IM formulations, such as aripiprazole, olanzapine, and ziprasidone, have been shown to be at least as effective as haloperidol for the treatment of acute agitation (in 2015, the short-acting injectable formulation of aripiprazole was discontinued in the United States independent of safety or efficacy issues⁵³). A review of RCTs examining IM SGAs for the treatment of agitation concluded that the number needed to treat for response compared with placebo was 5 for aripiprazole, 3 for olanzapine, and 3 for ziprasidone.⁵⁴ In terms of safety, a meta-analysis of studies examining IM medications for agitation confirmed that the risk of acute EPS, including dystonia, akathisia, and parkinsonism, is significantly lower with SGAs compared with haloperidol.⁵⁵ An RCT comparing IM ziprasidone with haloperidol found equivalently modest effects on QTc prolongation.⁵⁶ Therefore, SGAs are an obvious and evidence-based option for replacing haloperidol as a treatment for acute agitation.

Unfortunately, for clinicians hoping to replace haloperidol within a multiple-medication IM “cocktail,” there have been no published controlled trials of SGAs combined with benzodiazepines. Although a short report indicated that aripiprazole and lorazepam are chemically compatible to be combined within a single injection,⁵⁷ the package insert for aripiprazole warns that “If parenteral benzodiazepine therapy is deemed necessary in addition to ABILIFY injection treatment, patients should be monitored for excessive sedation and for orthostatic hypotension.”⁵⁸ The package insert for olanzapine likewise lists the combination of lorazepam and olanzapine as a drug interaction that can potentiate sedation, and the manufacturer issued specific warnings about parenteral combination.^59,60 A single published case of significant hypotension with combined IM olanzapine and lorazepam,⁶⁰ together with the fact that IM olanzapine can cause hypotension by itself,⁶¹ has discouraged the coadministration of these medications. Nonetheless, the combination is used in some emergency settings, with several retro­spective studies failing to provide evidence of hypotension or respiratory depression as adverse effects.^62-64

Droperidol. Droperidol was formerly a popular choice for managing acute agitation, with evidence from RCTs that droperidol, 5 mg, can improve symptoms significantly faster than either haloperidol, 5 mg, or lorazepam, 2 mg, and is absorbed just as rapidly whether administered IV or IM.^65-67 However, a 2001 FDA black-box warning about QTc prolongation included recommendations that a screening electrocardiogram should be obtained before administering droperidol. This action greatly curtailed the use of droperidol, and for some time, it was not marketed or available in the United States.

Over the past decade, however, droperidol has returned to the US market⁶⁸ and its IV and IM usage has been revitalized for managing patients with agitation within or en route to the ED. Studies have demonstrated droperidol efficacy comparable to midazolam, ziprasidone, or olanzapine, as well as effectiveness as an IV adjunct to midazolam.^69-71 In contrast to the FDA black-box warning, retrospective studies and RCTs of both IV and IM droperidol suggest that QTc prolongation and torsades de pointes are rare events that do not occur any more frequently than they do with haloperidol, even at doses >10 mg.^72,73 However, in studies involving patients with drug intoxication and treatment with multiple medications, oversedation to the point of needing rescue intervention was reported. In an emergency setting where these issues are relatively easily managed, such risks may be better tolerated than in psychiatric settings.

With earlier studies examining the use of droperidol in an acute psychiatric setting that reported a more rapid onset of action than haloperidol,^65-67 a 2016 Cochrane review concluded that there was high-quality evidence to support droperidol’s use for psychosis-induced agitation.⁷⁴ However, a 2015 RCT comparing IM droperidol, 10 mg, to haloperidol, 10 mg, found equivalent efficacy and response times (with maximal response occurring within 2 hours) and concluded that droperidol had no advantage over haloperidol.⁷⁵ Because none of the clinical trials that evaluated droperidol have included assessments for EPS, its risk of akathisia remains uncertain.

Continue to: Ketamine

Ketamine. In recent years, ketamine has been used to treat acute agitation within or en route to the ED. Preliminary observational studies support ketamine’s efficacy when administered via IV or IM routes,⁷⁶ with more rapid symptomatic improvement compared with haloperidol, lorazepam, or midazolam alone.⁷⁷ Reported adverse effects of ketamine include dissociation, psychotic exacerbation, and respiratory depression,⁷⁶ although 1 small naturalistic study found no evidence of exacerbation of psychotic or other psychiatric symptoms.⁷⁸ An ongoing RCT is comparing IM ketamine, 5 mg/kg, to combined IM haloperidol, 5 mg, and midazolam, 5 mg.⁷⁹ Although various ketamine formulations are increasingly being used in psychiatry, active psychosis is generally regarded as a contraindication. It is premature to recommend parenteral ketamine administration for agitation within most psychiatric settings until more research on safety has been completed.

Haloperidol, or something else? Practical considerations

Consider the following factors when deciding whether to use haloperidol or one of its alternatives:

Limitations of the evidence. Modern clinical trials requiring informed consent often do not include the kind of severe agitation that clinicians encounter in acute psychiatric, emergency, or forensic settings. In addition, standard interventions, such as 3-medication haloperidol “cocktails,” have not been evaluated in clinical trials. Clinicians are therefore often in the dark about optimal evidence-based practices.

Treatment goals. Psychiatric agitation has many causes, with a range of severity that warrants a commensurate range of responses. Protocols for managing acute agitation should include graded interventions that begin with nonpharmacologic interventions and voluntary oral medications, and move to involuntary IM medications when necessary.

While treatment guidelines clearly recommend against IM medications as “chemical restraint” with a goal of sedating a patient until he/she is unconscious,^3,51 such outcomes are nonetheless often sought by staff who are concerned about the risk of injuries during a behavioral emergency. In such instances, the risks of violence towards patients and staff may outweigh concerns about adverse effects in a risk-benefit analysis. Consequently, clinicians may be prone to “skip over” graded interventions because they assume they “won’t work” in favor of administering involuntary multiple-medication haloperidol “cocktails” despite risks of excess sedation, EPS, and cardiotoxicity. Treatment settings should critically evaluate such biased preferences, with a goal of developing tailored, evidence-based strategies that maximize benefits while minimizing excess sedation and other untoward adverse effects, with an eye towards promoting better overall patient care and reducing length of stay.^42,43,80

Continue to: Limitations of available medications

Limitations of available medications. There is no perfect medication for the management of acute agitation. Evidence indicates that pharmacologic options take 15 minutes to several hours to resolve acute agitation, even potentially more rapid-acting medications such as midazolam and droperidol. This is well beyond most clinicians’ desired window for response time in a behavioral emergency. Multiple-medication “cocktails” may be used with the hope of hastening response time, but may not achieve this goal at the expense of increasing the risk of adverse effects and the likelihood that a patient will remain sedated for a prolonged time. In the real world, this often means that by the time a psychiatrist comes to evaluate a patient who has been given emergency medications, the patient cannot be aroused for an interview. Ideally, medications would calm an agitated patient rapidly, without excess or prolonged sedation.⁸⁰ Less-sedating SGAs, such as ziprasidone, might have this potential, but can sometimes be perceived as ineffective.

Avoiding akathisia. Akathisia’s potential to worsen and be mistaken for agitation makes it an especially concerning, if underappreciated, adverse effect of haloperidol that is often not adequately assessed in clinical trials or practice. In light of evidence that akathisia can occur in nearly half of patients receiving a single 5 mg-dose of haloperidol, it is difficult to justify the use of this medication for agitation when equally effective options exist with a lower risk of EPS.

While haloperidol-induced akathisia could in theory be mitigated by adding anticholinergic medications or benzodiazepines, previous studies have found that such strategies have limited effectiveness compared to “gold standard” treatment with propranolol.^28,81,82 Furthermore, the half-lives of anticholinergic medications, such as benztropine or diphenhydramine, are significantly shorter than that of a single dose of haloperidol, which can be as long as 37 hours.⁸³ Therefore, akathisia and other EPS could emerge or worsen several hours or even days after receiving an IM haloperidol “cocktail” as the shorter-acting medications wear off. Akathisia is best minimized by avoiding FGAs, such as haloperidol, when treating acute agitation.

Promoting adherence. Although haloperidol is often recommended for acute agitation in patients with schizophrenia or bipolar disorder on the basis that it would treat the underlying condition, many patients who receive IM medications for acute agitation are already prescribed standing doses of oral medication, which increases the risk of cumulative toxicity. In addition, receiving a medication likely to cause acute EPS that is ranked near the bottom of patient preferences may erode the potential for a therapeutic alliance and hamper longer-term antipsychotic medication adherence.

Time for a change

For nearly half a century, haloperidol has been a “gold standard” intervention for IM control in patients with agitation. However, given its potential to produce adverse effects, including a significant risk of akathisia that can worsen agitation, along with the availability of newer pharmacologic options that are at least as effective (Table 1, and Table 2), haloperidol should be retired as a first-line medication for the treatment of agitation. Clinicians would benefit from RCTs investigating the safety and efficacy of novel interventions including frequently-used, but untested medication combinations, as well as nonpharmacologic interventions.

Continue to: Bottom Line

Although there is no perfect IM medication to treat acute agitation, haloperidol’s higher risk of adverse effects relative to newer alternatives suggest that it should no longer be considered a first-line intervention.

Related Resources

• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 1: onset of efficacy. J Emerg Med. 2018;54(3):364-374.
• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 2: safety. J Emerg Med. 2018;54(4): 522-532.

Drug Brand Names

Aripiprazole • Abilify
Benztropine • Cogentin
Chlorpromazine • Thorazine
Diphenhydramine • Benadryl
Droperidol • Inapsine
Haloperidol • Haldol
Ketamine • Ketalar
Lorazepam • Ativan
Midazolam • Versed
Olanzapine • Zyprexa
Prochlorperazine • Compazine
Promethazine • Phenergan
Propranolol • Inderal, Pronol
Ziprasidone • Geodon

For more than half a century, haloperidol has been used as a first-line medication for psychiatric agitation constituting a “behavioral emergency” when a patient cannot or will not take oral medication. Today, haloperidol is most commonly administered as an IM injection along with an anticholinergic medication to minimize extrapyramidal symptoms (EPS) and a benzodiazepine for additional sedation. The multiple-medication “cocktail” is often referred to by double-entendre nicknames, such as “B-52” or “5250” (ie, haloperidol, 5 mg; lorazepam, 2 mg; and diphenhydramine, 50 mg). In this article, I discuss whether haloperidol, a first-generation antipsychotic (FGA) medication developed in 1958, still deserves to be the IM “gold standard” for managing emergency psychiatric agitation.

Earlier evidence of haloperidol’s efficacy

The initial “discovery” of antipsychotic medications was made in 1951 based on the inadvertent observation that chlorpromazine had the potential to calm surgical patients with autonomic activation. This calming effect, described as “désintéressment” (meaning a kind of “indifference to the world”),¹ resulted in a new class of medications replacing barbiturates and bromides as go-to options to achieve “rapid tranquilization” of psychiatric agitation.² Although the ability of antipsychotic medications to gradually reduce positive symptoms, such as delusions and hallucinations, has been attributed to dopamine (D2) antagonism, their more immediate sedating and anti-agitation effects are the result of broader effects as histamine (H1) and alpha-1 adrenergic antagonists.

In the 1970s, haloperidol emerged as a first-line option to manage agitation due to its IM and IV availability, as well as its relative lack of sedation and orthostasis compared with low-potency D2 antagonists such as chlorpromazine. However, haloperidol was observed to have a significant risk of acute EPS, including dystonic reactions.² From the 1970s to the 1990s, numerous prospective clinical trials of haloperidol for the treatment of acute psychotic agitation, including several randomized controlled trials (RCTs) comparing haloperidol to lorazepam, were conducted.³ The design and outcomes of the haloperidol vs lorazepam RCTs were fairly consistent^4-7:

• adult participants with acute agitation and a variety of psychiatric diagnoses, for whom informed consent often was waived due to agitation severity
• randomization to either IM haloperidol, 5 mg, or IM lorazepam, 2 mg, administered every 30 minutes until agitation resolved
• behavioral outcomes measured over several hours using various rating scales, without consistent assessment of EPS
• equivalent efficacy of haloperidol and lorazepam, with symptom resolution usually achieved after 1 to 2 doses (in 30 to 60 minutes), but sometimes longer
• anticholinergic “rescue” allowed for EPS, but not administered prophylactically
• EPS, including dystonia and akathisia, were significantly more frequent with haloperidol compared with lorazepam.⁸

In recognition of the greater risk of EPS with haloperidol compared with lorazepam, and the fact that most study participants were already taking standing doses of antipsychotic medications, some researchers have recommended using benzodiazepines alone as the optimal treatment for agitation.^4,9 A 2012 Cochrane review concluded that the involuntary use of haloperidol alone “could be considered unethical.”^10,11 However, other studies that examined the combination of haloperidol and lorazepam compared with either medication alone found that the combination of the 2 medications was associated with a more rapid resolution of symptoms, which suggests a superior synergistic effect.^6,7,12 By the late 1990s, combined haloperidol and lorazepam, often mixed within a single injection, became the most common strategy to achieve rapid tranquilization in the psychiatric emergency setting.¹³ However, while the combination has been justified as a way to reduce the antipsychotic medication dose and EPS risk,² few studies have compared combinations containing <5 mg of haloperidol. As a result, the apparent superiority of combined haloperidol and lorazepam compared with either medication alone may be a simple cumulative dose effect rather than true synergism. It is also important to note that adding lorazepam to haloperidol does not mitigate the risk of EPS such as dystonia in the absence of anticholinergic medication.⁸ To date, however, there have been no clinical trials investigating the efficacy of IM haloperidol, lorazepam, and benztropine or diphenhydramine given together.

Newer RCTs tell a different story

With the availability of second-generation antipsychotics (SGAs) in IM formulations, clinical trials over the past 2 decades have focused on comparing SGAs with haloperidol alone as the “gold standard” control for acute agitation. Compared with previous trials of haloperidol vs lorazepam, these clinical trials of SGAs vs haloperidol included^8,14-22:

• Study participants who signed informed consent (and were likely less agitated)
• IM haloperidol doses typically >5 mg (eg, 6.5 to 10 mg).

As with studies comparing lorazepam with haloperidol, the results of these RCTs revealed that IM aripiprazole, olanzapine, and ziprasidone were at least as effective as IM haloperidol, with haloperidol having a significantly increased risk of akathisia, dystonia, and other EPS.^8,14-22 The greater EPS risk of haloperidol is not surprising given the use of comparison doses up to 10 mg.

An updated 2017 Cochrane review of haloperidol for psychosis-induced aggression or agitation concluded that⁹:

• haloperidol is an effective intervention, although the evidence is “weak”
• significant treatment effects may take as long as 1 to 2 hours following multiple IM injections
• in contrast to SGAs, treatment with haloperidol carries a significant risk of EPS
• adding a benzodiazepine “does not have strong evidence of benefit and carries risk of additional harm.”

Continue to: Haloperidol's well-known toxicity

Haloperidol has been associated with numerous adverse effects:

Akathisia and other acute EPS. Treatment with even a single dose of IM haloperidol can result in acute EPS, including dystonia and akathisia. At best, such adverse effects are subjectively troubling and unpleasant; at worst, akathisia can exacerbate and be mistaken for agitation, leading to administration of more medication²³ and the possible development of suicidal or violent behavior.^24-25 In the studies reviewed above, the overall rate of EPS was as high as 21% after treatment with haloperidol,¹⁶ with parkinsonism occurring in up to 17% of patients,¹⁹ dystonia in up to 11%,⁷ and akathisia in up to 10%.¹⁵ However, because specific EPS were assessed inconsistently, and sometimes not at all, the rate of akathisia—arguably the most relevant and counter-therapeutic adverse effect related to agitation—remains unclear.

In another study that specifically assessed for akathisia in patients treated with haloperidol, up to 40% experienced akathisia 6 hours after a single oral dose of 5 mg.²⁶ Even a single dose of IV prochlorperazine, another dopamine-antagonist routinely used to treat nausea in the emergency department (ED), has been reported to cause akathisia in up to 44% of patients.²⁷ Such results suggest that when akathisia is carefully assessed, the rate with even brief FGA exposure may approach nearly half of treated patients. Because akathisia is typically dose-related, and considering that many patients receiving IM haloperidol may receive multiple injections in addition to standing doses of oral medications, akathisia may be underrecognized in patients who are agitated, with a much greater risk than is generally presumed.

Although anticholinergic medications or benzodiazepinesare often administered as part of a haloperidol “cocktail,” these medications often do not adequately resolve emergent akathisia.^26,28 No clinical trials of IM haloperidol combined with benztropine or diphenhydramine have been published, but several studies suggest that combining haloperidol with promethazine—a phenothiazine with strong antihistaminergic and anticholinergic activity, but only weak antidopaminergic activity—can decrease the risk of dystonia relative to haloperidol alone.^8,22,29,30 However, there have also been reports of promethazine causing dystonia.^31,32 In addition, 1 trial of IM haloperidol, 2.5 mg, combined with promethazine reported that 74% of patients still had at least 1 form of EPS.³⁰ Because the clinical trials of haloperidol with promethazine did not specifically assess for akathisia, promethazine’s ability to decrease the risk of akathisia remains unknown.

Cardiotoxicity. Although low-potency antipsychotic medications such as chlorpromazine are more sedating than haloperidol, the latter is preferred as an IM antipsychotic medication for agitation because of its lower risk of hypotension.² In terms of cardiac effects, all antipsychotic medications carry a risk of QTc prolongation, with possible progression to the potentially lethal arrhythmia torsades de pointes as a result of interference with cardiac potassium channels.³³ In 2007, the FDA added a “black-box” warning about this risk for haloperidol, in the wake of a disproportionately high number of reported cases associated with IV administration, sometimes even after a single dose.³⁴

Continue to: Although there is no direct evidence...

Although there is no direct evidence that the cardiac risks associated with IV haloperidol apply to IM administration, epidemiologic studies indicate that oral haloperidol carries an elevated risk of ventricular arrhythmia and sudden cardiac death,^35,36 with 1 study reporting greater risk compared with other SGAs.³⁷ Haloperidol, whether administered orally or IM, may therefore be an especially poor choice for patients with agitation who are at risk for arrhythmia, including those with relevant medical comorbidities or delirium.³⁴

Neuronal cell death. Several lines of research evidence have demonstrated that haloperidol can cause cellular injury or death in neuronal tissue in a dose-dependent fashion through a variety of mechanisms.³⁸ By contrast, SGAs have been shown to have neuroprotective effects.³⁹ While these findings have mostly come from studies conducted in animals or in vitro human tumor cell lines, some researchers have nonetheless called for haloperidol to be banned, noting that if its neurotoxic effects were more widely known, “we would realize what a travesty it is to use [such] a brain-unfriendly drug.”⁴⁰

Several reasonable alternatives

Echoing the earlier Cochrane review of haloperidol for psychosis-induced aggression or agitation,¹⁰ a 2017 update concluded, “If no other alternative exists, sole use of intramuscular haloperidol could be life-saving. Where additional drugs are available, sole use of haloperidol for extreme emergency could be considered unethical.”⁹

What then are reasonable alternatives to replace IM haloperidol for agitation? Clinicians should consider the following nonpharmacologic and pharmacologic interventions:

Nonpharmacologic interventions. Several behavioral interventions have been demonstrated to be effective for managing acute agitation, including verbal de-escalation, enhanced “programming” on the inpatient units, and the judicious use of seclusion.^41-43 While such interventions may demand additional staff or resources, they have the potential to lower long-term costs, reduce injuries to patients and staff, and improve the quality of care.⁴³ The use of IM haloperidol as a form of “chemical restraint” does not represent standard-of-care treatment,³ and from an ethical perspective, should never be implemented punitively or to compensate for substandard care in the form of inadequate staffing or staff training.

Continue to: Benzodiazepines

Benzodiazepines. Lorazepam offers an attractive alternative to haloperidol without the risk of EPS.^2,4,8 However, lorazepam alone may be perceived as less efficacious than a haloperidol “cocktail” because it represents less overall medication. Some evidence has suggested that lorazepam, 4 mg, might be the most appropriate dose, although it has only rarely been studied in clinical trials of acute agitation.³ Midazolam is another IM benzodiazepine alternative to IM haloperidol that has been shown to achieve more rapid sedation than either haloperidol or lorazepam,^44,45 although it can cause substantial anterograde amnesia and also has an FDA black-box warning for respiratory depression associated with IV administration.

Respiratory depression is frequently cited as an argument against using lorazepam for agitation, as if the therapeutic window is extremely narrow with ineffectiveness at 2 mg, but potential lethality beyond that dose. In fact, serious respiratory depression with lorazepam is unlikely in the absence of chronic obstructive pulmonary disease (COPD), obstructive sleep apnea, or concomitant alcohol or other sedative use.⁴⁶ Case reports have documented therapeutic lorazepam dosing of 2 to 4 mg every 2 hours up to 20 to 30 mg/d in patients with manic agitation.⁴⁷ Even in patients with COPD, significant respiratory depression tends not to occur at doses <8 mg.⁴⁸ A more evidence-based concern about lorazepam dosing is that 2 mg might be ineffective in patients with established tolerance. For example, 1 report described a patient in acute alcohol withdrawal who required dosing lorazepam to 1,600 mg within 24 hours.⁴⁹ Collectively, these reports suggest that lorazepam has a much wider therapeutic window than is typically perceived, and that dosing with 3 to 4 mg IM is a reasonable option for agitation when 2 mg is likely to be inadequate.

Paradoxical disinhibition is another concern that might prevent benzodiazepines from being used alone as a first-line intervention for emergency treatment of agitation. However, similar to respiratory depression, this adverse event is relatively rare and tends to occur in children and geriatric patients, individuals intoxicated with alcohol or other sedatives, and patients with brain injury, developmental delay, or dementia.^23,46 Although exacerbation of aggression has not been demonstrated in the RCTs examining benzodiazepines for agitation reviewed above, based on other research, some clinicians have expressed concerns about the potential for benzodiazepines to exacerbate aggression in patients with impulse control disorders and a history of violent behavior.⁵⁰

The 2005 Expert Consensus Panel for Behavioral Emergencies⁵¹ recommended the use of lorazepam alone over haloperidol for agitation for patients for whom the diagnosis is unknown or includes the following:

• stimulant intoxication
• personality disorder
• comorbid obesity
• comorbid cardiac arrhythmia
• a history of akathisia and other EPS
• a history of amenorrhea/galactorrhea
• a history of seizures.

In surveys, patients have ranked lorazepam as the preferred medication for emergency agitation, whereas haloperidol was ranked as one of the least-preferred options.^51,52

Continue to: Second-generation antipsychotics

Second-generation antipsychotics. The SGAs available in IM formulations, such as aripiprazole, olanzapine, and ziprasidone, have been shown to be at least as effective as haloperidol for the treatment of acute agitation (in 2015, the short-acting injectable formulation of aripiprazole was discontinued in the United States independent of safety or efficacy issues⁵³). A review of RCTs examining IM SGAs for the treatment of agitation concluded that the number needed to treat for response compared with placebo was 5 for aripiprazole, 3 for olanzapine, and 3 for ziprasidone.⁵⁴ In terms of safety, a meta-analysis of studies examining IM medications for agitation confirmed that the risk of acute EPS, including dystonia, akathisia, and parkinsonism, is significantly lower with SGAs compared with haloperidol.⁵⁵ An RCT comparing IM ziprasidone with haloperidol found equivalently modest effects on QTc prolongation.⁵⁶ Therefore, SGAs are an obvious and evidence-based option for replacing haloperidol as a treatment for acute agitation.

Unfortunately, for clinicians hoping to replace haloperidol within a multiple-medication IM “cocktail,” there have been no published controlled trials of SGAs combined with benzodiazepines. Although a short report indicated that aripiprazole and lorazepam are chemically compatible to be combined within a single injection,⁵⁷ the package insert for aripiprazole warns that “If parenteral benzodiazepine therapy is deemed necessary in addition to ABILIFY injection treatment, patients should be monitored for excessive sedation and for orthostatic hypotension.”⁵⁸ The package insert for olanzapine likewise lists the combination of lorazepam and olanzapine as a drug interaction that can potentiate sedation, and the manufacturer issued specific warnings about parenteral combination.^59,60 A single published case of significant hypotension with combined IM olanzapine and lorazepam,⁶⁰ together with the fact that IM olanzapine can cause hypotension by itself,⁶¹ has discouraged the coadministration of these medications. Nonetheless, the combination is used in some emergency settings, with several retro­spective studies failing to provide evidence of hypotension or respiratory depression as adverse effects.^62-64

Droperidol. Droperidol was formerly a popular choice for managing acute agitation, with evidence from RCTs that droperidol, 5 mg, can improve symptoms significantly faster than either haloperidol, 5 mg, or lorazepam, 2 mg, and is absorbed just as rapidly whether administered IV or IM.^65-67 However, a 2001 FDA black-box warning about QTc prolongation included recommendations that a screening electrocardiogram should be obtained before administering droperidol. This action greatly curtailed the use of droperidol, and for some time, it was not marketed or available in the United States.

Over the past decade, however, droperidol has returned to the US market⁶⁸ and its IV and IM usage has been revitalized for managing patients with agitation within or en route to the ED. Studies have demonstrated droperidol efficacy comparable to midazolam, ziprasidone, or olanzapine, as well as effectiveness as an IV adjunct to midazolam.^69-71 In contrast to the FDA black-box warning, retrospective studies and RCTs of both IV and IM droperidol suggest that QTc prolongation and torsades de pointes are rare events that do not occur any more frequently than they do with haloperidol, even at doses >10 mg.^72,73 However, in studies involving patients with drug intoxication and treatment with multiple medications, oversedation to the point of needing rescue intervention was reported. In an emergency setting where these issues are relatively easily managed, such risks may be better tolerated than in psychiatric settings.

With earlier studies examining the use of droperidol in an acute psychiatric setting that reported a more rapid onset of action than haloperidol,^65-67 a 2016 Cochrane review concluded that there was high-quality evidence to support droperidol’s use for psychosis-induced agitation.⁷⁴ However, a 2015 RCT comparing IM droperidol, 10 mg, to haloperidol, 10 mg, found equivalent efficacy and response times (with maximal response occurring within 2 hours) and concluded that droperidol had no advantage over haloperidol.⁷⁵ Because none of the clinical trials that evaluated droperidol have included assessments for EPS, its risk of akathisia remains uncertain.

Continue to: Ketamine

Ketamine. In recent years, ketamine has been used to treat acute agitation within or en route to the ED. Preliminary observational studies support ketamine’s efficacy when administered via IV or IM routes,⁷⁶ with more rapid symptomatic improvement compared with haloperidol, lorazepam, or midazolam alone.⁷⁷ Reported adverse effects of ketamine include dissociation, psychotic exacerbation, and respiratory depression,⁷⁶ although 1 small naturalistic study found no evidence of exacerbation of psychotic or other psychiatric symptoms.⁷⁸ An ongoing RCT is comparing IM ketamine, 5 mg/kg, to combined IM haloperidol, 5 mg, and midazolam, 5 mg.⁷⁹ Although various ketamine formulations are increasingly being used in psychiatry, active psychosis is generally regarded as a contraindication. It is premature to recommend parenteral ketamine administration for agitation within most psychiatric settings until more research on safety has been completed.

Haloperidol, or something else? Practical considerations

Consider the following factors when deciding whether to use haloperidol or one of its alternatives:

Limitations of the evidence. Modern clinical trials requiring informed consent often do not include the kind of severe agitation that clinicians encounter in acute psychiatric, emergency, or forensic settings. In addition, standard interventions, such as 3-medication haloperidol “cocktails,” have not been evaluated in clinical trials. Clinicians are therefore often in the dark about optimal evidence-based practices.

Treatment goals. Psychiatric agitation has many causes, with a range of severity that warrants a commensurate range of responses. Protocols for managing acute agitation should include graded interventions that begin with nonpharmacologic interventions and voluntary oral medications, and move to involuntary IM medications when necessary.

While treatment guidelines clearly recommend against IM medications as “chemical restraint” with a goal of sedating a patient until he/she is unconscious,^3,51 such outcomes are nonetheless often sought by staff who are concerned about the risk of injuries during a behavioral emergency. In such instances, the risks of violence towards patients and staff may outweigh concerns about adverse effects in a risk-benefit analysis. Consequently, clinicians may be prone to “skip over” graded interventions because they assume they “won’t work” in favor of administering involuntary multiple-medication haloperidol “cocktails” despite risks of excess sedation, EPS, and cardiotoxicity. Treatment settings should critically evaluate such biased preferences, with a goal of developing tailored, evidence-based strategies that maximize benefits while minimizing excess sedation and other untoward adverse effects, with an eye towards promoting better overall patient care and reducing length of stay.^42,43,80

Continue to: Limitations of available medications

Limitations of available medications. There is no perfect medication for the management of acute agitation. Evidence indicates that pharmacologic options take 15 minutes to several hours to resolve acute agitation, even potentially more rapid-acting medications such as midazolam and droperidol. This is well beyond most clinicians’ desired window for response time in a behavioral emergency. Multiple-medication “cocktails” may be used with the hope of hastening response time, but may not achieve this goal at the expense of increasing the risk of adverse effects and the likelihood that a patient will remain sedated for a prolonged time. In the real world, this often means that by the time a psychiatrist comes to evaluate a patient who has been given emergency medications, the patient cannot be aroused for an interview. Ideally, medications would calm an agitated patient rapidly, without excess or prolonged sedation.⁸⁰ Less-sedating SGAs, such as ziprasidone, might have this potential, but can sometimes be perceived as ineffective.

Avoiding akathisia. Akathisia’s potential to worsen and be mistaken for agitation makes it an especially concerning, if underappreciated, adverse effect of haloperidol that is often not adequately assessed in clinical trials or practice. In light of evidence that akathisia can occur in nearly half of patients receiving a single 5 mg-dose of haloperidol, it is difficult to justify the use of this medication for agitation when equally effective options exist with a lower risk of EPS.

While haloperidol-induced akathisia could in theory be mitigated by adding anticholinergic medications or benzodiazepines, previous studies have found that such strategies have limited effectiveness compared to “gold standard” treatment with propranolol.^28,81,82 Furthermore, the half-lives of anticholinergic medications, such as benztropine or diphenhydramine, are significantly shorter than that of a single dose of haloperidol, which can be as long as 37 hours.⁸³ Therefore, akathisia and other EPS could emerge or worsen several hours or even days after receiving an IM haloperidol “cocktail” as the shorter-acting medications wear off. Akathisia is best minimized by avoiding FGAs, such as haloperidol, when treating acute agitation.

Promoting adherence. Although haloperidol is often recommended for acute agitation in patients with schizophrenia or bipolar disorder on the basis that it would treat the underlying condition, many patients who receive IM medications for acute agitation are already prescribed standing doses of oral medication, which increases the risk of cumulative toxicity. In addition, receiving a medication likely to cause acute EPS that is ranked near the bottom of patient preferences may erode the potential for a therapeutic alliance and hamper longer-term antipsychotic medication adherence.

Time for a change

For nearly half a century, haloperidol has been a “gold standard” intervention for IM control in patients with agitation. However, given its potential to produce adverse effects, including a significant risk of akathisia that can worsen agitation, along with the availability of newer pharmacologic options that are at least as effective (Table 1, and Table 2), haloperidol should be retired as a first-line medication for the treatment of agitation. Clinicians would benefit from RCTs investigating the safety and efficacy of novel interventions including frequently-used, but untested medication combinations, as well as nonpharmacologic interventions.

Continue to: Bottom Line

Although there is no perfect IM medication to treat acute agitation, haloperidol’s higher risk of adverse effects relative to newer alternatives suggest that it should no longer be considered a first-line intervention.

Related Resources

• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 1: onset of efficacy. J Emerg Med. 2018;54(3):364-374.
• Zun LS. Evidence-based review of pharmacotherapy for acute agitation. Part 2: safety. J Emerg Med. 2018;54(4): 522-532.

Drug Brand Names

Aripiprazole • Abilify
Benztropine • Cogentin
Chlorpromazine • Thorazine
Diphenhydramine • Benadryl
Droperidol • Inapsine
Haloperidol • Haldol
Ketamine • Ketalar
Lorazepam • Ativan
Midazolam • Versed
Olanzapine • Zyprexa
Prochlorperazine • Compazine
Promethazine • Phenergan
Propranolol • Inderal, Pronol
Ziprasidone • Geodon

Over the past few decades, it has become increasingly clear that cannabis use can increase the risk of developing a psychotic disorder and worsen the course of existing schizophrenia in a dose-dependent fashion.^1-3 Beyond psychosis, although many patients with mental illness use cannabis for recreational purposes or as purported “self-medication,” currently available evidence suggests that marijuana is more likely to represent a harm than a benefit for psychi­atric disorders⁴ (Box^4-8). Our current state of knowledge therefore suggests that psychiatrists should caution their patients against using cannabis and prioritize interventions to reduce or discontinue use, especially among those with psychotic disorders.

Box

Despite that conclusion, because cannabis is classified as a Schedule I drug by the US Drug Enforcement Agency, clinical research investigating the risks and benefits of cannabis has been limited. It therefore remains possible that cannabis, or individual cannabinoids such as cannabidiol (CBD), may yet find a therapeutic niche in psychiatry. This article reviews evidence on CBD for the treatment of schizophrenia.

Cannabinergic drugs as potential antipsychotics

Although the bulk of evidence indicates a harmful effect of cannabis in individuals with or at risk for psychosis, there have been a few published cases of schizophrenia improving with dronabinol, an FDA-approved, synthetic form of delta-9-tetrahydrocannabinol (THC).^9,10 THC is the constituent of cannabis that produces euphoric effects. These provocative findings have not been replicated in controlled clinical trials, but suggest at least the theoretical possibility of idiosyncratic benefits from THC for some individuals within the psychotic spectrum.

Still, given that most available evidence supports that THC has a harmful effect on psychosis and psychosis risk, researchers have instead performed randomized controlled trials (RCTs) to investigate a possible therapeutic role for medications that oppose the agonist effects of THC at cannabinoid type 1 (CB1) receptors. To date, 2 RCTs comparing rimonabant, a CB1 inverse agonist, with placebo (PLB) in patients with schizophrenia have failed to demonstrate any benefit for psychotic symptoms or cognitive deficits.^11,12 A third trial examining rimonabant for people diagnosed with schizophrenia who were overweight found significant benefits for anxiety and depressive symptoms, but none for positive symptoms or the primary outcome of weight loss.¹³ While these results are discouraging, the role of THC in precipitating psychosis suggests that novel agents opposing the actions of THC on the cannabinoid system could have antipsychotic properties.¹⁴

Cannabidiol: An antipsychotic medication?

In contrast to THC, CBD has minimal euphorigenic properties and has recently been heralded in the popular press as a “miracle drug” with benefits for medical and psychiatric disorders alike.¹⁵ It has even been speculated that it could become a popular food supplement.¹⁶ In 2018, the FDA gave full approval to a pharmaceutically manufactured form of CBD (brand name: Epidiolex) as a novel treatment for 2 rare and severe forms of pediatric epilepsy, Lennox-Gastaut syndrome and Dravet syndrome,¹⁷ based on RCTs supporting its efficacy for these often refractory and life-threatening conditions.^18-20

In psychiatry, there have not yet been enough robust clinical studies to support broad therapeutic claims for CBD as a treatment for any mental disorder.²¹ However, there is growing evidence that CBD has potential as an antipsychotic medication. In 1995, the first case report was published describing the efficacy of CBD, 1,500 mg/d, as standalone therapy in a single individual with schizophrenia.²² In 2006, the same research group followed up with a case series in which only 1 out of 3 patients with treatment-refractory schizophrenia improved with flexible dosing of CBD to a maximum dose of 1,280 mg/d.²³

There have been 3 published RCTs exploring the efficacy of CBD in schizophrenia (Table^24-26). The first study, published in 2012, included 39 adults with schizophrenia who were randomized to 800 mg/d of CBD or amisulpride (AMS), a second-generation antipsychotic that is popular in Europe but is not available in the United States.²⁴ Over 4 weeks of randomized treatment, CBD resulted in as much improvement in overall symptoms and positive symptoms as AMS, and improvement of negative symptoms was significantly greater with CBD. Compared with patients treated with antipsychotic medication, patients who were treated with CBD had fewer extrapyramidal symptoms, less weight gain, and less prolactin elevation. This initial trial suggests that CBD might be as efficacious in schizophrenia as antipsychotic medication, without its burdensome adverse effects. However, this is the only RCT of CBD monotherapy published to date.

Continue to: Two other recently published RCTs...

Two other recently published RCTs compared CBD with PLB as add-on therapy to antipsychotics. McGuire et al²⁵ compared CBD, 1,000 mg/d, to PLB over 6 weeks in 88 patients with schizophrenia. Positive symptom improvement was statistically greater with CBD than with PLB, although the magnitude of clinical change was modest (using the Positive and Negative Syndrome Scale [PANSS] positive symptom subscale: −3.2 points for CBD vs −1.7 points for PLB). Changes in PANSS total score and subscales for general and negative symptoms were not significantly different between treatment groups. There was also no significant difference in overall change in neurocognitive symptoms, although post-hoc analysis revealed significantly greater improvement in motor speed for patients treated with CBD. More than twice the number of patients treated with CBD were rated as “much improved” by the Clinical Global Impressions scale compared with patients treated with PLB, but this was not a statistically significant finding, and most patients experienced only “minimal” or “no improvement.” In terms of adverse events, there were no significant differences between patients in the CBD and PLB groups. Although this study is technically “positive” for CBD and suggests minimal adverse effects, it is not clear whether the statistically significant positive symptom improvements (+1.5 PANSS points for CBD over PLB) were clinically significant.

The most recently published placebo-controlled RCT of CBD as add-on therapy to antipsychotic medication included 36 patients with schizophrenia treated over 6 weeks.²⁶ In this study, there was no benefit of CBD, 600 mg/d, on any PANSS score outcome (total, general, positive, or negative symptoms). For the primary outcome of the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery, there were no significant drug × time effects, and post-hoc analyses showed that only patients treated with PLB improved with time. Sedation was more common among patients treated with CBD compared with PLB.

Making sense of the data

There have been mixed results from the few case reports and 3 RCTs of patients with schizophrenia who were treated with CBD. How can we resolve these disparate findings? A few possible interpretations of the data that warrant clarification through additional research include:

Dosing. In the first case report with positive results, CBD was dosed at 1,500 mg/d,²² whereas in the subsequent case series with mixed results, the maximum allowable dose of CBD was 1,280 mg/d.²³ Likewise, in the RCTs, positive results were found when CBD was dosed at 800 to 1,000 mg/d,^24,25 but not at 600 mg/d.²⁶ The efficacy of CBD for schizophrenia might depend on higher doses.

Treatment resistance. In the second case series in which only 1 out of 3 patients responded to treatment with CBD,²³ the patients had demonstrated previous nonresponse to at least 2 first-generation antipsychotics (FGAs) and risperidone, 6 mg/d. In the RCTs, all patients were antipsychotic-responsive.^24-26 Cannabidiol may not be as effective for patients with treatment-refractory schizophrenia as it is for patients with schizophrenia who respond to antipsychotics.

Continue to: Clinical stability

Answering questions about CBD

Cannabidiol is becoming increasingly popular for its purported health benefits. The mixed results of the few studies published on CBD for schizophrenia place clinicians in a difficult position when attempting to answer questions about how cannabinoids might fit into treatment of patients with psychosis. Consider the following:

Is cannabis helpful for patients with schizophrenia? No. Aside from the few case reports suggesting that FDA-approved THC (dronabinol) can improve symptoms in some patients,^9,10 most of the evidence from anecdotal reports and both experimental and observational studies indicate that cannabis, THC, and synthetic cannabinoids have a harmful effect in patients with or at risk for psychosis.^1-3

If you are considering recommending some form of cannabis to patients with schizophrenia, what kind should you recommend? Recommending or encouraging cannabis use for patients with psychosis is ill-advised. Although certain types of cannabis might contain more THC (eg, Cannabis indica vs Cannabis sativa) or variable amounts of CBD, in general the amount of CBD in whole leaf cannabis is minimal, with the ratio of THC to CBD increasingly significantly over the past decade.^3,27 Most forms of cannabis should therefore be avoided by individuals with or at risk for psychotic disorders.

Continue to: What about CBD oil and other CBD products sold in dispensaries?

What about CBD oil and other CBD products sold in dispensaries? Cannabidiol is increasingly available in various forms based on its ability to be designated as a legal hemp product (containing <0.3% THC) at the federal level or as a cannabinoid in states where cannabis is legal. However, several studies have now shown that cannabis products sold online or in dispensaries are often labeled inaccurately, with both under- and over-reporting of THC and CBD content.^28-30 Some CBD products have been found to have almost no CBD at all.^29,30 The unreliability of product labeling makes it difficult to predict the effects of CBD products that are not subject to FDA purity standards for medications or dietary supplements. It also raises questions about the sources of CBD and the reliability of dosing in the studies discussed above.

Why might CBD work as an antipsychotic? Although CBD has minimal affinity for cannabinoid receptors, it appears to act as a partial agonist of dopamine D2 receptors and an agonist at 5-HT1A receptors, with overall effects that decrease mesolimbic dopamine activity.^31,32 In addition, CBD increases the availability of the endogenous cannabinoid anandamide, which may have antipsychotic properties.^14,33

Now that the FDA has approved CBD manufactured by a pharmaceutical company, should it be prescribed “off-label” for patients with schizophrenia? This is the “million dollar question,” with insufficient evidence to provide a clear answer. It should now be possible to prescribe FDA-approved CBD for off-label purposes, including the treatment of schizophrenia and other psychiatric disorders. No doubt, some clinicians are already doing so. This will predictably yield more anecdotal evidence about efficacy and adverse effects in the future, but there is not yet adequate evidence to support an FDA indication for CBD in schizophrenia. Additional studies of CBD for schizophrenia are ongoing.

Bottom Line

Cannabidiol (CBD) is becoming increasingly popular based on its purported health benefits, but the evidence supporting a therapeutic role in psychiatry is preliminary at best. Although CBD is now available by prescription as an FDA-approved drug for the treatment of 2 rare forms of epilepsy, its benefits in patients with schizophrenia are uncertain based on mixed results in clinical trials.

Related Resources

• Clinicaltrials.gov. Studies of “cannabidiol” and “schizophrenia.” U.S. National Library of Medicine. https://clinicaltrials.gov/ct2/results?cond=Schizophrenia&term=cannabidiol.
• Grinspoon P. Cannabidiol (CBD) – what we know and what we don’t. Harvard Health Blog. https://www.health.harvard.edu/blog/cannabidiol-cbd-what-we-know-and-what-wedont-2018082414476. Published August 24, 2018.

Drug Brand Names

Cannabidiol • Epidiolex
Dronabinol • Marinol
Risperidone • Risperdal

Over the past few decades, it has become increasingly clear that cannabis use can increase the risk of developing a psychotic disorder and worsen the course of existing schizophrenia in a dose-dependent fashion.^1-3 Beyond psychosis, although many patients with mental illness use cannabis for recreational purposes or as purported “self-medication,” currently available evidence suggests that marijuana is more likely to represent a harm than a benefit for psychi­atric disorders⁴ (Box^4-8). Our current state of knowledge therefore suggests that psychiatrists should caution their patients against using cannabis and prioritize interventions to reduce or discontinue use, especially among those with psychotic disorders.

Box

Despite that conclusion, because cannabis is classified as a Schedule I drug by the US Drug Enforcement Agency, clinical research investigating the risks and benefits of cannabis has been limited. It therefore remains possible that cannabis, or individual cannabinoids such as cannabidiol (CBD), may yet find a therapeutic niche in psychiatry. This article reviews evidence on CBD for the treatment of schizophrenia.

Cannabinergic drugs as potential antipsychotics

Although the bulk of evidence indicates a harmful effect of cannabis in individuals with or at risk for psychosis, there have been a few published cases of schizophrenia improving with dronabinol, an FDA-approved, synthetic form of delta-9-tetrahydrocannabinol (THC).^9,10 THC is the constituent of cannabis that produces euphoric effects. These provocative findings have not been replicated in controlled clinical trials, but suggest at least the theoretical possibility of idiosyncratic benefits from THC for some individuals within the psychotic spectrum.

Still, given that most available evidence supports that THC has a harmful effect on psychosis and psychosis risk, researchers have instead performed randomized controlled trials (RCTs) to investigate a possible therapeutic role for medications that oppose the agonist effects of THC at cannabinoid type 1 (CB1) receptors. To date, 2 RCTs comparing rimonabant, a CB1 inverse agonist, with placebo (PLB) in patients with schizophrenia have failed to demonstrate any benefit for psychotic symptoms or cognitive deficits.^11,12 A third trial examining rimonabant for people diagnosed with schizophrenia who were overweight found significant benefits for anxiety and depressive symptoms, but none for positive symptoms or the primary outcome of weight loss.¹³ While these results are discouraging, the role of THC in precipitating psychosis suggests that novel agents opposing the actions of THC on the cannabinoid system could have antipsychotic properties.¹⁴

Cannabidiol: An antipsychotic medication?

In contrast to THC, CBD has minimal euphorigenic properties and has recently been heralded in the popular press as a “miracle drug” with benefits for medical and psychiatric disorders alike.¹⁵ It has even been speculated that it could become a popular food supplement.¹⁶ In 2018, the FDA gave full approval to a pharmaceutically manufactured form of CBD (brand name: Epidiolex) as a novel treatment for 2 rare and severe forms of pediatric epilepsy, Lennox-Gastaut syndrome and Dravet syndrome,¹⁷ based on RCTs supporting its efficacy for these often refractory and life-threatening conditions.^18-20

In psychiatry, there have not yet been enough robust clinical studies to support broad therapeutic claims for CBD as a treatment for any mental disorder.²¹ However, there is growing evidence that CBD has potential as an antipsychotic medication. In 1995, the first case report was published describing the efficacy of CBD, 1,500 mg/d, as standalone therapy in a single individual with schizophrenia.²² In 2006, the same research group followed up with a case series in which only 1 out of 3 patients with treatment-refractory schizophrenia improved with flexible dosing of CBD to a maximum dose of 1,280 mg/d.²³

There have been 3 published RCTs exploring the efficacy of CBD in schizophrenia (Table^24-26). The first study, published in 2012, included 39 adults with schizophrenia who were randomized to 800 mg/d of CBD or amisulpride (AMS), a second-generation antipsychotic that is popular in Europe but is not available in the United States.²⁴ Over 4 weeks of randomized treatment, CBD resulted in as much improvement in overall symptoms and positive symptoms as AMS, and improvement of negative symptoms was significantly greater with CBD. Compared with patients treated with antipsychotic medication, patients who were treated with CBD had fewer extrapyramidal symptoms, less weight gain, and less prolactin elevation. This initial trial suggests that CBD might be as efficacious in schizophrenia as antipsychotic medication, without its burdensome adverse effects. However, this is the only RCT of CBD monotherapy published to date.

Continue to: Two other recently published RCTs...

Two other recently published RCTs compared CBD with PLB as add-on therapy to antipsychotics. McGuire et al²⁵ compared CBD, 1,000 mg/d, to PLB over 6 weeks in 88 patients with schizophrenia. Positive symptom improvement was statistically greater with CBD than with PLB, although the magnitude of clinical change was modest (using the Positive and Negative Syndrome Scale [PANSS] positive symptom subscale: −3.2 points for CBD vs −1.7 points for PLB). Changes in PANSS total score and subscales for general and negative symptoms were not significantly different between treatment groups. There was also no significant difference in overall change in neurocognitive symptoms, although post-hoc analysis revealed significantly greater improvement in motor speed for patients treated with CBD. More than twice the number of patients treated with CBD were rated as “much improved” by the Clinical Global Impressions scale compared with patients treated with PLB, but this was not a statistically significant finding, and most patients experienced only “minimal” or “no improvement.” In terms of adverse events, there were no significant differences between patients in the CBD and PLB groups. Although this study is technically “positive” for CBD and suggests minimal adverse effects, it is not clear whether the statistically significant positive symptom improvements (+1.5 PANSS points for CBD over PLB) were clinically significant.

The most recently published placebo-controlled RCT of CBD as add-on therapy to antipsychotic medication included 36 patients with schizophrenia treated over 6 weeks.²⁶ In this study, there was no benefit of CBD, 600 mg/d, on any PANSS score outcome (total, general, positive, or negative symptoms). For the primary outcome of the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery, there were no significant drug × time effects, and post-hoc analyses showed that only patients treated with PLB improved with time. Sedation was more common among patients treated with CBD compared with PLB.

Making sense of the data

There have been mixed results from the few case reports and 3 RCTs of patients with schizophrenia who were treated with CBD. How can we resolve these disparate findings? A few possible interpretations of the data that warrant clarification through additional research include:

Dosing. In the first case report with positive results, CBD was dosed at 1,500 mg/d,²² whereas in the subsequent case series with mixed results, the maximum allowable dose of CBD was 1,280 mg/d.²³ Likewise, in the RCTs, positive results were found when CBD was dosed at 800 to 1,000 mg/d,^24,25 but not at 600 mg/d.²⁶ The efficacy of CBD for schizophrenia might depend on higher doses.

Treatment resistance. In the second case series in which only 1 out of 3 patients responded to treatment with CBD,²³ the patients had demonstrated previous nonresponse to at least 2 first-generation antipsychotics (FGAs) and risperidone, 6 mg/d. In the RCTs, all patients were antipsychotic-responsive.^24-26 Cannabidiol may not be as effective for patients with treatment-refractory schizophrenia as it is for patients with schizophrenia who respond to antipsychotics.

Continue to: Clinical stability

Answering questions about CBD

Cannabidiol is becoming increasingly popular for its purported health benefits. The mixed results of the few studies published on CBD for schizophrenia place clinicians in a difficult position when attempting to answer questions about how cannabinoids might fit into treatment of patients with psychosis. Consider the following:

Is cannabis helpful for patients with schizophrenia? No. Aside from the few case reports suggesting that FDA-approved THC (dronabinol) can improve symptoms in some patients,^9,10 most of the evidence from anecdotal reports and both experimental and observational studies indicate that cannabis, THC, and synthetic cannabinoids have a harmful effect in patients with or at risk for psychosis.^1-3

If you are considering recommending some form of cannabis to patients with schizophrenia, what kind should you recommend? Recommending or encouraging cannabis use for patients with psychosis is ill-advised. Although certain types of cannabis might contain more THC (eg, Cannabis indica vs Cannabis sativa) or variable amounts of CBD, in general the amount of CBD in whole leaf cannabis is minimal, with the ratio of THC to CBD increasingly significantly over the past decade.^3,27 Most forms of cannabis should therefore be avoided by individuals with or at risk for psychotic disorders.

Continue to: What about CBD oil and other CBD products sold in dispensaries?

What about CBD oil and other CBD products sold in dispensaries? Cannabidiol is increasingly available in various forms based on its ability to be designated as a legal hemp product (containing <0.3% THC) at the federal level or as a cannabinoid in states where cannabis is legal. However, several studies have now shown that cannabis products sold online or in dispensaries are often labeled inaccurately, with both under- and over-reporting of THC and CBD content.^28-30 Some CBD products have been found to have almost no CBD at all.^29,30 The unreliability of product labeling makes it difficult to predict the effects of CBD products that are not subject to FDA purity standards for medications or dietary supplements. It also raises questions about the sources of CBD and the reliability of dosing in the studies discussed above.

Why might CBD work as an antipsychotic? Although CBD has minimal affinity for cannabinoid receptors, it appears to act as a partial agonist of dopamine D2 receptors and an agonist at 5-HT1A receptors, with overall effects that decrease mesolimbic dopamine activity.^31,32 In addition, CBD increases the availability of the endogenous cannabinoid anandamide, which may have antipsychotic properties.^14,33

Now that the FDA has approved CBD manufactured by a pharmaceutical company, should it be prescribed “off-label” for patients with schizophrenia? This is the “million dollar question,” with insufficient evidence to provide a clear answer. It should now be possible to prescribe FDA-approved CBD for off-label purposes, including the treatment of schizophrenia and other psychiatric disorders. No doubt, some clinicians are already doing so. This will predictably yield more anecdotal evidence about efficacy and adverse effects in the future, but there is not yet adequate evidence to support an FDA indication for CBD in schizophrenia. Additional studies of CBD for schizophrenia are ongoing.

Bottom Line

Cannabidiol (CBD) is becoming increasingly popular based on its purported health benefits, but the evidence supporting a therapeutic role in psychiatry is preliminary at best. Although CBD is now available by prescription as an FDA-approved drug for the treatment of 2 rare forms of epilepsy, its benefits in patients with schizophrenia are uncertain based on mixed results in clinical trials.

Related Resources

• Clinicaltrials.gov. Studies of “cannabidiol” and “schizophrenia.” U.S. National Library of Medicine. https://clinicaltrials.gov/ct2/results?cond=Schizophrenia&term=cannabidiol.
• Grinspoon P. Cannabidiol (CBD) – what we know and what we don’t. Harvard Health Blog. https://www.health.harvard.edu/blog/cannabidiol-cbd-what-we-know-and-what-wedont-2018082414476. Published August 24, 2018.

Drug Brand Names

Cannabidiol • Epidiolex
Dronabinol • Marinol
Risperidone • Risperdal

Over the past few decades, it has become increasingly clear that cannabis use can increase the risk of developing a psychotic disorder and worsen the course of existing schizophrenia in a dose-dependent fashion.^1-3 Beyond psychosis, although many patients with mental illness use cannabis for recreational purposes or as purported “self-medication,” currently available evidence suggests that marijuana is more likely to represent a harm than a benefit for psychi­atric disorders⁴ (Box^4-8). Our current state of knowledge therefore suggests that psychiatrists should caution their patients against using cannabis and prioritize interventions to reduce or discontinue use, especially among those with psychotic disorders.

Box

Despite that conclusion, because cannabis is classified as a Schedule I drug by the US Drug Enforcement Agency, clinical research investigating the risks and benefits of cannabis has been limited. It therefore remains possible that cannabis, or individual cannabinoids such as cannabidiol (CBD), may yet find a therapeutic niche in psychiatry. This article reviews evidence on CBD for the treatment of schizophrenia.

Cannabinergic drugs as potential antipsychotics

Although the bulk of evidence indicates a harmful effect of cannabis in individuals with or at risk for psychosis, there have been a few published cases of schizophrenia improving with dronabinol, an FDA-approved, synthetic form of delta-9-tetrahydrocannabinol (THC).^9,10 THC is the constituent of cannabis that produces euphoric effects. These provocative findings have not been replicated in controlled clinical trials, but suggest at least the theoretical possibility of idiosyncratic benefits from THC for some individuals within the psychotic spectrum.

Still, given that most available evidence supports that THC has a harmful effect on psychosis and psychosis risk, researchers have instead performed randomized controlled trials (RCTs) to investigate a possible therapeutic role for medications that oppose the agonist effects of THC at cannabinoid type 1 (CB1) receptors. To date, 2 RCTs comparing rimonabant, a CB1 inverse agonist, with placebo (PLB) in patients with schizophrenia have failed to demonstrate any benefit for psychotic symptoms or cognitive deficits.^11,12 A third trial examining rimonabant for people diagnosed with schizophrenia who were overweight found significant benefits for anxiety and depressive symptoms, but none for positive symptoms or the primary outcome of weight loss.¹³ While these results are discouraging, the role of THC in precipitating psychosis suggests that novel agents opposing the actions of THC on the cannabinoid system could have antipsychotic properties.¹⁴

Cannabidiol: An antipsychotic medication?

In contrast to THC, CBD has minimal euphorigenic properties and has recently been heralded in the popular press as a “miracle drug” with benefits for medical and psychiatric disorders alike.¹⁵ It has even been speculated that it could become a popular food supplement.¹⁶ In 2018, the FDA gave full approval to a pharmaceutically manufactured form of CBD (brand name: Epidiolex) as a novel treatment for 2 rare and severe forms of pediatric epilepsy, Lennox-Gastaut syndrome and Dravet syndrome,¹⁷ based on RCTs supporting its efficacy for these often refractory and life-threatening conditions.^18-20

In psychiatry, there have not yet been enough robust clinical studies to support broad therapeutic claims for CBD as a treatment for any mental disorder.²¹ However, there is growing evidence that CBD has potential as an antipsychotic medication. In 1995, the first case report was published describing the efficacy of CBD, 1,500 mg/d, as standalone therapy in a single individual with schizophrenia.²² In 2006, the same research group followed up with a case series in which only 1 out of 3 patients with treatment-refractory schizophrenia improved with flexible dosing of CBD to a maximum dose of 1,280 mg/d.²³

There have been 3 published RCTs exploring the efficacy of CBD in schizophrenia (Table^24-26). The first study, published in 2012, included 39 adults with schizophrenia who were randomized to 800 mg/d of CBD or amisulpride (AMS), a second-generation antipsychotic that is popular in Europe but is not available in the United States.²⁴ Over 4 weeks of randomized treatment, CBD resulted in as much improvement in overall symptoms and positive symptoms as AMS, and improvement of negative symptoms was significantly greater with CBD. Compared with patients treated with antipsychotic medication, patients who were treated with CBD had fewer extrapyramidal symptoms, less weight gain, and less prolactin elevation. This initial trial suggests that CBD might be as efficacious in schizophrenia as antipsychotic medication, without its burdensome adverse effects. However, this is the only RCT of CBD monotherapy published to date.

Continue to: Two other recently published RCTs...

Two other recently published RCTs compared CBD with PLB as add-on therapy to antipsychotics. McGuire et al²⁵ compared CBD, 1,000 mg/d, to PLB over 6 weeks in 88 patients with schizophrenia. Positive symptom improvement was statistically greater with CBD than with PLB, although the magnitude of clinical change was modest (using the Positive and Negative Syndrome Scale [PANSS] positive symptom subscale: −3.2 points for CBD vs −1.7 points for PLB). Changes in PANSS total score and subscales for general and negative symptoms were not significantly different between treatment groups. There was also no significant difference in overall change in neurocognitive symptoms, although post-hoc analysis revealed significantly greater improvement in motor speed for patients treated with CBD. More than twice the number of patients treated with CBD were rated as “much improved” by the Clinical Global Impressions scale compared with patients treated with PLB, but this was not a statistically significant finding, and most patients experienced only “minimal” or “no improvement.” In terms of adverse events, there were no significant differences between patients in the CBD and PLB groups. Although this study is technically “positive” for CBD and suggests minimal adverse effects, it is not clear whether the statistically significant positive symptom improvements (+1.5 PANSS points for CBD over PLB) were clinically significant.

The most recently published placebo-controlled RCT of CBD as add-on therapy to antipsychotic medication included 36 patients with schizophrenia treated over 6 weeks.²⁶ In this study, there was no benefit of CBD, 600 mg/d, on any PANSS score outcome (total, general, positive, or negative symptoms). For the primary outcome of the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery, there were no significant drug × time effects, and post-hoc analyses showed that only patients treated with PLB improved with time. Sedation was more common among patients treated with CBD compared with PLB.

Making sense of the data

There have been mixed results from the few case reports and 3 RCTs of patients with schizophrenia who were treated with CBD. How can we resolve these disparate findings? A few possible interpretations of the data that warrant clarification through additional research include:

Dosing. In the first case report with positive results, CBD was dosed at 1,500 mg/d,²² whereas in the subsequent case series with mixed results, the maximum allowable dose of CBD was 1,280 mg/d.²³ Likewise, in the RCTs, positive results were found when CBD was dosed at 800 to 1,000 mg/d,^24,25 but not at 600 mg/d.²⁶ The efficacy of CBD for schizophrenia might depend on higher doses.

Treatment resistance. In the second case series in which only 1 out of 3 patients responded to treatment with CBD,²³ the patients had demonstrated previous nonresponse to at least 2 first-generation antipsychotics (FGAs) and risperidone, 6 mg/d. In the RCTs, all patients were antipsychotic-responsive.^24-26 Cannabidiol may not be as effective for patients with treatment-refractory schizophrenia as it is for patients with schizophrenia who respond to antipsychotics.

Continue to: Clinical stability

Answering questions about CBD

Cannabidiol is becoming increasingly popular for its purported health benefits. The mixed results of the few studies published on CBD for schizophrenia place clinicians in a difficult position when attempting to answer questions about how cannabinoids might fit into treatment of patients with psychosis. Consider the following:

Is cannabis helpful for patients with schizophrenia? No. Aside from the few case reports suggesting that FDA-approved THC (dronabinol) can improve symptoms in some patients,^9,10 most of the evidence from anecdotal reports and both experimental and observational studies indicate that cannabis, THC, and synthetic cannabinoids have a harmful effect in patients with or at risk for psychosis.^1-3

If you are considering recommending some form of cannabis to patients with schizophrenia, what kind should you recommend? Recommending or encouraging cannabis use for patients with psychosis is ill-advised. Although certain types of cannabis might contain more THC (eg, Cannabis indica vs Cannabis sativa) or variable amounts of CBD, in general the amount of CBD in whole leaf cannabis is minimal, with the ratio of THC to CBD increasingly significantly over the past decade.^3,27 Most forms of cannabis should therefore be avoided by individuals with or at risk for psychotic disorders.

Continue to: What about CBD oil and other CBD products sold in dispensaries?

What about CBD oil and other CBD products sold in dispensaries? Cannabidiol is increasingly available in various forms based on its ability to be designated as a legal hemp product (containing <0.3% THC) at the federal level or as a cannabinoid in states where cannabis is legal. However, several studies have now shown that cannabis products sold online or in dispensaries are often labeled inaccurately, with both under- and over-reporting of THC and CBD content.^28-30 Some CBD products have been found to have almost no CBD at all.^29,30 The unreliability of product labeling makes it difficult to predict the effects of CBD products that are not subject to FDA purity standards for medications or dietary supplements. It also raises questions about the sources of CBD and the reliability of dosing in the studies discussed above.

Why might CBD work as an antipsychotic? Although CBD has minimal affinity for cannabinoid receptors, it appears to act as a partial agonist of dopamine D2 receptors and an agonist at 5-HT1A receptors, with overall effects that decrease mesolimbic dopamine activity.^31,32 In addition, CBD increases the availability of the endogenous cannabinoid anandamide, which may have antipsychotic properties.^14,33

Now that the FDA has approved CBD manufactured by a pharmaceutical company, should it be prescribed “off-label” for patients with schizophrenia? This is the “million dollar question,” with insufficient evidence to provide a clear answer. It should now be possible to prescribe FDA-approved CBD for off-label purposes, including the treatment of schizophrenia and other psychiatric disorders. No doubt, some clinicians are already doing so. This will predictably yield more anecdotal evidence about efficacy and adverse effects in the future, but there is not yet adequate evidence to support an FDA indication for CBD in schizophrenia. Additional studies of CBD for schizophrenia are ongoing.

Bottom Line

Cannabidiol (CBD) is becoming increasingly popular based on its purported health benefits, but the evidence supporting a therapeutic role in psychiatry is preliminary at best. Although CBD is now available by prescription as an FDA-approved drug for the treatment of 2 rare forms of epilepsy, its benefits in patients with schizophrenia are uncertain based on mixed results in clinical trials.

Related Resources

• Clinicaltrials.gov. Studies of “cannabidiol” and “schizophrenia.” U.S. National Library of Medicine. https://clinicaltrials.gov/ct2/results?cond=Schizophrenia&term=cannabidiol.
• Grinspoon P. Cannabidiol (CBD) – what we know and what we don’t. Harvard Health Blog. https://www.health.harvard.edu/blog/cannabidiol-cbd-what-we-know-and-what-wedont-2018082414476. Published August 24, 2018.

Drug Brand Names

Cannabidiol • Epidiolex
Dronabinol • Marinol
Risperidone • Risperdal

While some classes of medications used to treat psychi­atric disorders, such as stimulants and benzodiazepines, are well-recognized as controlled substances and drugs of abuse, clinicians may be less familiar with the potential misuse/abuse of other psychiatric medications. This article reviews the evidence related to the misuse/abuse of anticholinergics, antidepressants, antipsychotics, and gabapentinoids.

The terms “misuse,” “abuse,” and “addiction” are used variably in the literature without standardized definitions. For this review, “misuse/abuse (M/A)” will be used to collectively describe self-administration that is recreational or otherwise inconsistent with legal or medical guidelines, unless a specific distinction is made. Whether or not the medications reviewed are truly “addictive” will be briefly discussed for each drug class, but the focus will be on clinically relevant aspects of M/A, including:

• excessive self-administration
• self-administration by non-oral routes
• co-administration with other drugs of abuse
• malingering of psychiatric symptoms to obtain prescriptions
• diversion for sale to third parties
• toxicity from overdose.

Anticholinergic medications

The first case describing the deliberate M/A of an anticholinergic medication for its euphoric effects was published in 1960.Further reportsfollowed in Europe before the M/A potential of prescription anticholinergic medications among psychiatric patients with an overdose syndrome characterized by atropinism and toxic psychosis was more widely recognized in the United States in the 1970s. Most reported cases of M/A to date have occurred among patients with psychiatric illness because anticholinergic medications, including trihexyphenidyl, benztropine, biperiden, procyclidine, and orphenadrine, were commonly prescribed for the management of first-generation and high dopamine D2-affinity antipsychotic-induced extrapyramidal symptoms (EPS). For example, one study of 234 consecutively hospitalized patients with schizophrenia noted an anticholinergic M/A incidence of 6.5%.¹

However, anticholinergic M/A is not limited to individuals with psychotic disorders. A UK study of 154 admissions to an inpatient unit specializing in behavioral disturbances found a 12-month trihexyphenidyl M/A incidence of 17%; the most common diagnosis among abusers was antisocial personality disorder.² Anticholinergic M/A has also been reported among patients with a primary diagnosis of substance use disorders (SUDs)³ as well as more indiscriminately in prison settings,⁴ with some inmates exchanging trihexyphenidyl as currency and using it recreationally by crushing it into powder and smoking it with tobacco.⁵ Others have noted that abusers sometimes take anticholinergics with alcohol in order to “potentiate” the effects of each substance.^6,7 Pullen et al⁸ described individuals with and without psychiatric illness who stole anticholinergic medications, purchased them from other patients, or bought them “on the street.” Malingering EPS in order to obtain anticholinergic medications has also been well documented.⁹ Clearly, anticholinergic M/A can occur in psychiatric and non-psychiatric populations, both within and outside of clinical settings. Although anticholinergic M/A appears to be less frequent in the United States now that second-generation antipsychotics (SGAs) are more frequently prescribed, M/A remains common in some settings outside of the United States.⁷

Among the various anticholinergic medications prescribed for EPS, trihexyphenidyl has been reported to have the greatest M/A potential, which has been attributed to its potency,¹⁰ its stimulating effects (whereas benztropine is more sedating),¹¹ and its former popularity among prescribers.⁸ Marken et al¹¹ published a review of 110 reports of M/A occurring in patients receiving anticholinergic medications as part of psychiatric treatment in which 69% of cases involved taking trihexyphenidyl 15 to 60 mg at a time (recommended dosing is 6 to 10 mg/d in divided doses).Most of these patients were prescribed anticholinergic medications for diagnostically appropriate reasons—only 7% were described as “true abusers” with no medical indication. Anticholinergic M/A was typically driven by a desire for euphoric and psychedelic/hallucinogenic effects, although in some cases, anticholinergic M/A was attributed to self-medication of EPS and depressive symptoms. These findings illustrate the blurred distinction between recreational use and perceived subjective benefit, and match those of a subsequent study of 50 psychiatric patients who reported anticholinergic M/A not only to “get high,” but to “decrease depression,” “increase energy,” and decrease antipsychotic adverse effects.¹² Once again, trihexyphenidyl was the most frequently misused anticholinergic in this sample.

Table 1^{2,3,7,8,10-15} outlines the subjective effects sought and experienced by anticholinergic abusers as well as potential toxic effects; there is the potential for overlap. Several authors have also described physiologic dependence with long-term trihexyphenidyl use, including tolerance and a withdrawal/abstinence syndrome.^7,16 In addition, there have been several reports of coma¹³ and death in the setting of intended suicide by overdose of anticholinergic medications.^14,15

Although anticholinergic M/A in the United States now appears to be less common, clinicians should remain aware of the M/A potential of anticholinergic medications prescribed for EPS. Management of M/A involves:

• detection
• reducing anticholinergic exposure by managing EPS with alternative strategies, such as switching or reducing the dose of the antipsychotic medication
• gradual tapering of anticholinergic medications to minimize withdrawal.¹¹

Continue to: Antidepressants

Haddad¹⁷ published a review of 21 English-language case reports from 1966 to 1998 describing antidepressant use in which individuals met DSM-IV criteria for substance dependence to the medication. An additional 14 cases of antidepressant M/A were excluded based on insufficient details to support a diagnosis of dependence. The 21 reported cases involved:

• tranylcypromine (a monoamine oxidase inhibitor [MAOI])
• amitriptyline (a tricyclic antidepressant [TCA])
• fluoxetine (a selective serotonin reuptake inhibitor [SSRI])
• amineptine (a TCA previously available in France but removed from the market in 1999 in part due to its abuse potential)
• nomifensine (a norepinephrine/dopamine reuptake inhibitor previously available in the United Kingdom but removed in 1986 due to hemolytic anemia).

In 95% of cases, the antidepressants were prescribed for treatment of an affective disorder but were abused for stimulant effects or the perceived ability to lift mood, cause euphoria or a “high,” or to improve functioning. Two-thirds of cases involved patients with preexisting substance misuse. Placing the case reports in the context of the millions of patients prescribed antidepressants during this period, Haddad concluded the “incidence of [antidepressant] addiction [is] so low as to be clinically irrelevant.”¹⁷

Despite this conclusion, Haddad singled out amineptine and tranylcypromine as antidepressants with some evidence of true addictive potential.^17,18 A more recent case series described 14 patients who met DSM-IV criteria for substance abuse of tertiary amine TCAs (which have strong anticholinergic activity) and concluded that “misuse of [TCAs] is more common than generally appreciated.”¹⁹ In keeping with that claim, a study of 54 outpatients taking unspecified antidepressants found that up to 15% met DSM-III-R criteria for substance dependence (for the antidepressant) in the past year, although that rate was much lower than the rate of benzodiazepine dependence (47%) in a comparative sample.²⁰ Finally, a comprehensive review by Evans and Sullivan²¹ found anecdotal reports published before 2014 that detailed misuse, abuse, and dependence with MAOIs, TCAs, fluoxetine, venlafaxine, bupropion, tianeptine, and amineptine. Taken together, existing evidence indicates that select individuals—typically those with other SUD comorbidity—sometimes misuse antidepressants in a way that suggests addiction.

Still, while it is well known that abrupt cessation of antidepressants can result in a discontinuation syndrome characterized by flu-like symptoms, nausea, and dizziness,²² physiologic withdrawal effects must be distinguished from historical definitions of substance “abuse” and the broader concept of psychological “addiction” or drug dependence^18,23 now incorporated into the DSM-5 definition of SUDs.²⁴ Indeed, although withdrawal symptoms were reported by more than half of those who took antidepressants and responded to a recent online survey,²⁵ evidence to support the existence of significant antidepressant tolerance, craving, or compulsive use is lacking.^17,18 Antidepressants as a class do not appear to be significantly rewarding or reinforcing and, on the contrary, discontinuation by patients is common in clinical practice.²⁶ The popular claim that some individuals taking antidepressants “can’t quit”²⁷ must also be disentangled from loss of therapeutic effects upon cessation.

Bupropion. A more convincing argument for antidepressant addiction can be made for bupropion, a weak norepinephrine and dopamine reuptake inhibitor with an otherwise unclear mechanism of action.²⁸ In 2002, the first report of recreational bupropion M/A described a 13-year-old girl who took 2,400 mg orally (recommended maximum dose is 450 mg/d in divided doses) after being told it would give her “a better high than amphetamine.”²⁹ This was followed in the same year by the first report of recreational M/A of bupropion via nasal insufflation (snorting), resulting in a seizure,³⁰ and in 2013 by the first published case of M/A by IV self-administration.³¹

Continue to: The M/A potential of bupropion...

The M/A potential of bupropion, most commonly via intranasal administration, is now broadly recognized based on several case reports describing desired effects that include a euphoric high and a stimulating “buzz” similar to that of cocaine or methamphetamine but less intense.^29-36 Among recreational users, bupropion tablets are referred to as “welbys,” “wellies,” “dubs,” or “barnies.”³⁷ Media coverage of a 2013 outbreak of bupropion M/A in Toronto detailed administration by snorting, smoking, and injection, and described bupropion as “poor man’s cocaine.”³⁸ Between 2003 and 2016, 2,232 cases of bupropion misuse/abuse/dependence adverse drug reactions were reported to the European Monitoring Agency.³⁷ A review of intentional bupropion M/A reported to US Poison Control Centers between 2000 to 2013 found 975 such cases, with the yearly number tripling between 2000 and 2012.³⁹ In this sample, nearly half (45%) of the users were age 13 to 19, and 76% of cases involved oral ingestion. In addition to bupropion M/A among younger people, individuals who misuse bupropion often include those with existing SUDs but limited access to illicit stimulants and those trying to evade detection by urine toxicology screening.³³ For example, widespread use and diversion has been well documented within correctional settings, and as a result, many facilities have removed bupropion from their formularies.^{21,28,33,34,40}

Beyond desired effects, the most common adverse events associated with bupropion M/A are listed in Table 2,^{28,30,32-34,36,39} along with their incidence based on cases brought to the attention of US Poison Control Centers.³⁹ With relatively little evidence of a significant bupropion withdrawal syndrome,³⁷ the argument in favor of modeling bupropion as a truly addictive drug is limited to anecdotal reports of cravings and compulsive self-administration³⁵ and pro-dopaminergic activity (reuptake inhibition) that might provide a mechanism for potential rewarding and reinforcing effects.⁴⁰ While early preclinical studies of bupropion failed to provide evidence of amphetamine-like abuse potential,^41,42 non-oral administration in amounts well beyond therapeutic dosing could account for euphoric effects and a greater risk of psychological dependence and addiction.^21,28,40

Bupropion also has an FDA indication as an aid to smoking cessation treatment, and the medication demonstrated early promise in the pharmacologic treatment of psycho­stimulant use disorders, with reported improvements in cravings and other SUD outcomes.^43-45 However, subsequent randomized controlled trials (RCTs) failed to demonstrate a clear therapeutic role for bupropion in the treatment of cocaine^46,47 and methamphetamine use disorders (although some secondary analyses suggest possible therapeutic effects among non-daily stimulant users who are able to maintain good adherence with bupropion).^48-51 Given these overall discouraging results, the additive seizure risk of bupropion use with concomitant psychostimulant use, and the potential for M/A and diversion of bupropion (particularly among those with existing SUDs), the use of bupropion for the off-label treatment of stimulant use disorders is not advised.

Antipsychotics

As dopamine antagonists, antipsychotics are typically considered to have low potential for rewarding or reinforcing effects. Indeed, misuse of antipsychotics was a rarity in the first-generation era, with only a few published reports of haloperidol M/A within a small cluster of naïve young people who developed acute EPS,⁵² and a report of diversion in a prison with the “sadistic” intent of inflicting dystonic reactions on others.⁵³ A more recent report described 2additional cases of M/A involving haloperidol and trifluoperazine.⁵⁴ Some authors have described occasional drug-seeking behavior for low-potency D2 blockers such as chlorpromazine, presumably based on their M/A as anticholinergic medications.⁵⁵

The potential for antipsychotic M/A has gained wider recognition since the advent of the SGAs. Three cases of prescription olanzapine M/A have been published to date. One involved a man who malingered manic symptoms to obtain olanzapine, taking ≥40 mg at a time (beyond his prescribed dose of 20 mg twice daily) to get a “buzz,” and combining it with alcohol and benzodiazepines for additive effects or to “come down” from cocaine.⁵⁶ This patient noted that olanzapine was “a popular drug at parties” and was bought, sold, or traded among users, and occasionally administered intravenously. Two other cases described women who self-administered olanzapine, 40 to 50 mg/d, for euphoric and anxiolytic effects.^57,58 James et al⁵⁹ detailed a sample of 28 adults who reported “non-medical use” of olanzapine for anxiolytic effects, as a sleep aid, or to “escape from worries.”

Continue to: Quetiapine

Quetiapine. In contrast to some reports of olanzapine M/A in which the line between M/A and “self-medication” was blurred, quetiapine has become a more convincing example of clear recreational antipsychotic M/A. Since the first report of oral and intranasal quetiapine M/A in the Los Angeles County Jail published in 2004,⁵⁵ subsequent cases have detailed other novel methods of recreational self-administration^60-68 (Table 3^55,60-68), and additional reports have been published in non-English language journals.^69,70 Collectively, these case reports have detailed that quetiapine is:

• misused for primary subjective effects as well as to mitigate the unpleasant effects of other drugs^60,67
• referred to as “quell,”“Q,” “Susie-Q,” “squirrel,” and “baby heroin”^55,71,72
• often obtained by malingering psychiatric symptoms^55,61,63,65
• diverted/sold with “street value” both within and outside of psychiatric facilities and correctional settings.^{55,60-62,67,68,73}

These anecdotal accounts of quetiapine M/A have since been corroborated on a larger scale based on several retrospective studies. Although early reports of quetiapine M/A occurring in correctional settings have resulted in formulary removal,^71,74 quetiapine M/A is by no means limited to forensic populations and is especially common among those with comorbid SUDs. A survey of 74 patients enrolled in a Canadian methadone program reported that nearly 60% had misused quetiapine at some point.⁷⁵ Among an Australian sample of 868 individuals with active IV drug abuse, 31% reported having misused quetiapine.⁷⁶ Finally, within a small sample of patients with SUDs admitted to a detoxification unit in New York City, 17% reported M/A of SGAs.⁷⁷ In this study, SGAs were often taken in conjunction with other drugs of abuse in order to “recover” from or “enhance” the effects of other drugs or to “experiment.” Quetiapine was by far the most frequently abused SGA, reported in 96% of the sample; the most frequently reported SGA/drug combinations were quetiapine/alcohol/opioids, quetiapine/cocaine, and quetiapine/opioids.

Looking more broadly at poison center data, reports to the US National Poison Data System (NPDS) from 2005 to 2011 included 3,116 cases of quetiapine abuse (37.5%, defined as intentional recreational use in order to obtain a “high”) or misuse (62.5%, defined as improper use or dosing for non-recreational purposes).⁷⁸ A more recent analysis of NPDS reports from 2003 to 2013 found 2,118 cases of quetiapine abuse, representing 61% of all cases of reported SGA abuse.⁷⁹ An analysis of the European Medicines Agency Adverse Drug Database yielded 18,112 reports of quetia­pine misuse, abuse, dependence, and withdrawal for quetiapine (from 2005 to 2016) compared with 4,178 for olanzapine (from 2004 to 2016).⁸⁰ These reports identified 368 fatalities associated with quetiapine.

The rate of quetiapine M/A appears to be increasing sharply. Reports of quetiapine M/A to poison centers in Australia increased nearly 7-fold from 2006 to 2016.⁸¹ Based on reports to the Drug Abuse Warning System, US emergency department visits for M/A of quetiapine increased from 19,195 in 2005 to 32,024 in 2011 (an average of 27,114 visits/year), with 75% of cases involving quetiapine taken in combination with other prescription drugs, alcohol, or illicit drugs.⁸² Consistent with poison center data, M/A was reported for other antipsychotics, but none nearly as frequently as for quetiapine.

With increasingly frequent quetiapine M/A, clinicians should be vigilant in monitoring for medical morbidity related to quetiapine and cumulative toxicity with other drugs. The most frequent adverse events associated with quetiapine M/A reported to US Poison Control Centers are presented in Table 4.^78,79

Continue to: Unlike bupropion...

Unlike bupropion, quetiapine’s dopamine antagonism makes it unlikely to be a truly addictive drug, although this mechanism of action could mediate an increase in concurrent psychostimulant use.⁸³ A few case reports have described a quetiapine discontinuation syndrome similar to that of antidepressants,^60,65,84-88 but withdrawal symptoms suggestive of physiologic dependence may be mediated by non-dopaminergic effects through histamine and serotonin receptors.^84,89 Evidence for quetiapine misuse being associated with craving and compulsive use is lacking, and true quetiapine addiction is probably rare.

Similar to bupropion, preliminary findings have suggested promise for quetiapine as a putative therapy for other SUDs.^90-93 However, subsequent RCTs have failed to demonstrate a therapeutic effect for alcohol and cocaine use disorders.^94-96 Given these negative results and the clear M/A potential of quetiapine, off-label use of quetiapine for the treatment of SUDs and psychiatric symptoms among those with SUDs must be considered judiciously, with an eye towards possible diversion and avoiding the substitution of one drug of abuse for another.

Gabapentinoids

In 1997, the first published case report of gabapentin M/A described a woman who self-administered her husband’s gabapentin to reduce cravings for and withdrawal from cocaine.⁹⁷ The authors highlighted the possible therapeutic benefit of gabapentin in this regard rather than raising concerns about diversion and M/A. By 2004, however, reports of recreational gabapentin M/A emerged among inmates incarcerated within Florida correctional facilities who self-administered intranasal gabapentin to achieve a “high” that was “reminiscent of prior effects from intranasal ingestion of cocaine powder.”⁹⁸ In 2007, a single case of gabapentin misuse up to 7,200 mg/d (recommended dosing is ≤3,600 mg/d) was reported, with documentation of both tolerance and withdrawal symptoms.⁹⁹ As of 2017, a total of 36 cases of gabapentin M/A and 19 cases of pregabalin M/A have been published.¹⁰⁰

In the past decade, anecdotal reports have given way to larger-scale epidemiologic data painting a clear picture of the now-widespread M/A of gabapentin and other gabapentinoids. For example, a study of online descriptions of gabapentin and pregabalin M/A from 2008 to 2010 documented:

• oral and IM use (gabapentin)
• IV and rectal (“plugging”) use (pregabalin)
• “parachuting” (emptying the contents of capsules for a larger dose) (pregabalin)
• euphoric, entactogenic, stimulant, calming/anxiolytic, and dissociative subjective effects (gabapentin/pregabalin)
• rapid development of tolerance to euphoric effects leading to self-administration of increasing doses (gabapentin/pregabalin)
• frequent co-administration with other drugs of abuse, including alcohol, benzodiazepines, cannabis, stimulants, opiates, hallucinogens, gamma-hydroxybutyrate, mephedrone, and Salvia divinorum (gabapentin/pregabalin)¹⁰¹

Several systematic reviews of both anecdotal reports and epidemiologic studies published in the past few years provide additional evidence of the above, such as:

• excessive dosing with self-administration
• intranasal and inhaled routes of administration
• diversion and “street value”
• greater M/A potential of pregabalin than gabapentin
• the presence of gabapentinoids in postmortem toxicology analyses, suggesting a role in overdose fatalities when combined with other drugs.^100,102,103

Continue to: The European Medicine Agency's EudraVigilance database...

The European Medicine Agency’s EudraVigilance database included 4,301 reports of gabapentin misuse, abuse, or dependence, and 7,639 such reports for pregabalin, from 2006 to 2015 (rising sharply after 2012), with 86 gabapentin-related and 27 pregabalin-related fatalities.¹⁰⁴ Data from the Drug Diversion Program of the Researched Abuse, Diversion, and Addiction-Related Surveillance System from 2002 to 2015 have likewise revealed that gabapentin diversion increased significantly in 2013.¹⁰⁵

While the prevalence of gabapentinoid M/A is not known, rates appear to be significantly lower than for traditional drugs of abuse such as cannabis, cocaine, 3,4-methylenedioxymethamphetamine (MDMA), and opioids.^106,107 However, gabapentin and pregabalin M/A appears to be increasingly common among individuals with SUDs and in particular among those with opioid use disorders (OUDs). For example, a 2015 report indicated that 15% of an adult cohort in Appalachian Kentucky with nonmedical use of diverted prescription opioids reported gabapentin M/A, an increase of nearly 3,000% since 2008.¹⁰⁸ Based on data from a US insurance enrollment and claims database, researchers found that the rate of gabapentin overuse among those also overusing opioids was 12% compared with only 2% for those using gabapentin alone.¹⁰⁹ It has also been reported that gabapentin is sometimes used as a “cutting agent” for heroin.¹¹⁰

Those who use gabapentinoids together with opioids report that gabapentin and pregabalin potentiate the euphoric effects of methadone¹¹¹ and endorse specific beliefs that pregabalin increases both the desired effects of heroin as well as negative effects such as “blackouts,” loss of control, and risk of overdose.¹¹² Indeed, sustained M/A of gabapentin and opioids together has been found to increase emergency department utilization, drug-related hospitalization, and respiratory depression.¹¹³ Based on a case-control study of opioid users in Canada, co-prescription of gabapentin and opioids was associated with a 50% increase in death from opioid-related causes compared with prescription of opioids alone.¹¹⁴

Case reports documenting tolerance, withdrawal, craving, and loss of control suggest a true addictive potential for gabapentinoids, but Bonnet and Sherbaum¹⁰⁰ concluded that while there is robust evidence of abusers “liking” gabapentin and pregabalin (eg, reward), evidence of “wanting” them (eg, psychological dependence) in the absence of other SUDs has been limited to only a few anecdotal reports with pregabalin. Accordingly, the risk of true addiction to gabapentinoids by those without preexisting SUDs appears to be low. Nonetheless, the M/A potential of both gabapentin and pregabalin is clear and in the context of a nationwide opioid epidemic, the increased morbidity/mortality risk related to combined use of gabapentinoids and opioids is both striking and concerning. Consequently, the state of Kentucky recently recognized the M/A potential of gabapentin by designating it a Schedule V controlled substance (pregabalin is already a Schedule V drug according to the US Drug Enforcement Agency),^103,113 and several other states now mandate the reporting of gabapentin prescriptions to prescription drug monitoring programs.¹¹⁵

Following a similar pattern to antidepressants and antipsychotics, a potential role for gabapentin in the treatment of cocaine use disorders was supported in preliminary studies,^116-118 but not in subsequent RCTs.^119-121 However, there is evidence from RCTs to support the use of gabapentin and pregabalin in the treatment of alcohol use disorders.^122-124 Gabapentin was also found to significantly reduce cannabis use and withdrawal symptoms in patients compared with placebo in an RCT of individuals with cannabis use disorders.¹²⁵ The perceived safety of gabapentinoids by clinicians, their subjective desirability by patients with SUDs, and efficacy data supporting a therapeutic role in SUDs must be balanced with recognition that approximately 80% of gabapentin prescriptions are written for off-label indications for which there is little supporting evidence,¹⁰⁹ such as low back pain.¹²⁶ Clinicians considering prescribing gabapentinoids to manage psychiatric symptoms, such as anxiety and insomnia, should carefully consider the risk of M/A and other potential morbidities, especially in the setting of SUDs and OUD in particular.

Continue to: Problematic, even if not addictive

Problematic, even if not addictive

It is sometimes claimed that “addiction” to psychiatric medications is not limited to stimulants and benzodiazepines.^27,127 Although anticholinergics, antidepressants, antipsychotics, and gabapentinoids can be drugs of abuse, with some users reporting physiologic withdrawal upon discontinuation, there is only limited evidence that the M/A of these psychiatric medications is associated with the characteristic features of a more complete definition of “addiction,” which may include:

• inability to consistently abstain
• impairment in behavioral control
• diminished recognition of significant problems associated with use
• a dysfunctional emotional response to chronic use.¹²⁸

Nonetheless, the literature documenting anticholinergic, antidepressant, antipsychotic, and gabapentinoid M/A includes several common features, including:

• initial reports among those with limited access to illicit drugs (eg, young people and incarcerated individuals) and subsequent spread to a wider population with more unconventional routes of administration
• use for recreational purposes and other subjective pseudo-therapeutic effects, often in combination with alcohol and illicit drugs
• greater M/A potential of certain medications within each of these drug classes (eg, trihexyphenidyl, bupropion, quetiapine)
• malingering psychiatric symptoms in order to obtain medications from prescribers and diversion for black market sale
• observations that medications might constitute therapy for SUDs that were not supported in subsequent RCTs (with the exception of gabapentin for alcohol and cannabis use disorders)
• increasing evidence of toxicity related to M/A, which suggests that prescription by clinicians has limited benefit and high risk for patients with SUDs.

Bottom Line

Some psychiatric medications are taken as drugs of abuse. Clinicians should be particularly aware of the misuse/abuse potential of anticholinergics, antidepressants, antipsychotics, and gabapentinoids, and use them cautiously, if at all, when treating patients with existing substance use disorders.

Related Resources

• Substance Abuse and Mental Health Services Administration. Prescription drug misuse and abuse. https://www.samhsa.gov/topics/prescription-drug-misuse-abuse.
• Substance Abuse and Mental Health Services Administration. Types of commonly misused or abused drugs. https://www.samhsa.gov/prescription-drug-misuse-abuse/types.
• National Institute on Drug Abuse. Misuse of prescription drugs. https://www.drugabuse.gov/publications/research-reports/misuse-prescription-drugs/summary.
• National Institute on Drug Abuse. New clinician screening tool available for substance use. https://www.drugabuse.gov/news-events/news-releases/2018/06/newclinician-screening-tool-available-substance-use.

Drug Brand Names

Amitriptyline • Elavil, Endep
Benztropine • Cogentin
Biperiden • Akineton
Bupropion • Wellbutrin, Zyban
Chlorpromazine • Thorzine
Fluoxetine • Prozac
Haloperidol • Haldol
Olanzapine • Zyprexa
Orphenadrine • Disipal, Norflex
Pregabalin • Lyrica, Lyrica CR
Procyclidine • Kemadrin
Quetiapine • Seroquel
Tianeptine • Coaxil, Stablon
Tranylcypromine • Parnate
Trifluoperazine • Stelazine
Trihexyphenidyl • Artane, Tremin
Venlafaxine • Effexor

While some classes of medications used to treat psychi­atric disorders, such as stimulants and benzodiazepines, are well-recognized as controlled substances and drugs of abuse, clinicians may be less familiar with the potential misuse/abuse of other psychiatric medications. This article reviews the evidence related to the misuse/abuse of anticholinergics, antidepressants, antipsychotics, and gabapentinoids.

The terms “misuse,” “abuse,” and “addiction” are used variably in the literature without standardized definitions. For this review, “misuse/abuse (M/A)” will be used to collectively describe self-administration that is recreational or otherwise inconsistent with legal or medical guidelines, unless a specific distinction is made. Whether or not the medications reviewed are truly “addictive” will be briefly discussed for each drug class, but the focus will be on clinically relevant aspects of M/A, including:

• excessive self-administration
• self-administration by non-oral routes
• co-administration with other drugs of abuse
• malingering of psychiatric symptoms to obtain prescriptions
• diversion for sale to third parties
• toxicity from overdose.

Anticholinergic medications

The first case describing the deliberate M/A of an anticholinergic medication for its euphoric effects was published in 1960.Further reportsfollowed in Europe before the M/A potential of prescription anticholinergic medications among psychiatric patients with an overdose syndrome characterized by atropinism and toxic psychosis was more widely recognized in the United States in the 1970s. Most reported cases of M/A to date have occurred among patients with psychiatric illness because anticholinergic medications, including trihexyphenidyl, benztropine, biperiden, procyclidine, and orphenadrine, were commonly prescribed for the management of first-generation and high dopamine D2-affinity antipsychotic-induced extrapyramidal symptoms (EPS). For example, one study of 234 consecutively hospitalized patients with schizophrenia noted an anticholinergic M/A incidence of 6.5%.¹

However, anticholinergic M/A is not limited to individuals with psychotic disorders. A UK study of 154 admissions to an inpatient unit specializing in behavioral disturbances found a 12-month trihexyphenidyl M/A incidence of 17%; the most common diagnosis among abusers was antisocial personality disorder.² Anticholinergic M/A has also been reported among patients with a primary diagnosis of substance use disorders (SUDs)³ as well as more indiscriminately in prison settings,⁴ with some inmates exchanging trihexyphenidyl as currency and using it recreationally by crushing it into powder and smoking it with tobacco.⁵ Others have noted that abusers sometimes take anticholinergics with alcohol in order to “potentiate” the effects of each substance.^6,7 Pullen et al⁸ described individuals with and without psychiatric illness who stole anticholinergic medications, purchased them from other patients, or bought them “on the street.” Malingering EPS in order to obtain anticholinergic medications has also been well documented.⁹ Clearly, anticholinergic M/A can occur in psychiatric and non-psychiatric populations, both within and outside of clinical settings. Although anticholinergic M/A appears to be less frequent in the United States now that second-generation antipsychotics (SGAs) are more frequently prescribed, M/A remains common in some settings outside of the United States.⁷

Among the various anticholinergic medications prescribed for EPS, trihexyphenidyl has been reported to have the greatest M/A potential, which has been attributed to its potency,¹⁰ its stimulating effects (whereas benztropine is more sedating),¹¹ and its former popularity among prescribers.⁸ Marken et al¹¹ published a review of 110 reports of M/A occurring in patients receiving anticholinergic medications as part of psychiatric treatment in which 69% of cases involved taking trihexyphenidyl 15 to 60 mg at a time (recommended dosing is 6 to 10 mg/d in divided doses).Most of these patients were prescribed anticholinergic medications for diagnostically appropriate reasons—only 7% were described as “true abusers” with no medical indication. Anticholinergic M/A was typically driven by a desire for euphoric and psychedelic/hallucinogenic effects, although in some cases, anticholinergic M/A was attributed to self-medication of EPS and depressive symptoms. These findings illustrate the blurred distinction between recreational use and perceived subjective benefit, and match those of a subsequent study of 50 psychiatric patients who reported anticholinergic M/A not only to “get high,” but to “decrease depression,” “increase energy,” and decrease antipsychotic adverse effects.¹² Once again, trihexyphenidyl was the most frequently misused anticholinergic in this sample.

Table 1^{2,3,7,8,10-15} outlines the subjective effects sought and experienced by anticholinergic abusers as well as potential toxic effects; there is the potential for overlap. Several authors have also described physiologic dependence with long-term trihexyphenidyl use, including tolerance and a withdrawal/abstinence syndrome.^7,16 In addition, there have been several reports of coma¹³ and death in the setting of intended suicide by overdose of anticholinergic medications.^14,15

Although anticholinergic M/A in the United States now appears to be less common, clinicians should remain aware of the M/A potential of anticholinergic medications prescribed for EPS. Management of M/A involves:

• detection
• reducing anticholinergic exposure by managing EPS with alternative strategies, such as switching or reducing the dose of the antipsychotic medication
• gradual tapering of anticholinergic medications to minimize withdrawal.¹¹

Continue to: Antidepressants

Haddad¹⁷ published a review of 21 English-language case reports from 1966 to 1998 describing antidepressant use in which individuals met DSM-IV criteria for substance dependence to the medication. An additional 14 cases of antidepressant M/A were excluded based on insufficient details to support a diagnosis of dependence. The 21 reported cases involved:

• tranylcypromine (a monoamine oxidase inhibitor [MAOI])
• amitriptyline (a tricyclic antidepressant [TCA])
• fluoxetine (a selective serotonin reuptake inhibitor [SSRI])
• amineptine (a TCA previously available in France but removed from the market in 1999 in part due to its abuse potential)
• nomifensine (a norepinephrine/dopamine reuptake inhibitor previously available in the United Kingdom but removed in 1986 due to hemolytic anemia).

In 95% of cases, the antidepressants were prescribed for treatment of an affective disorder but were abused for stimulant effects or the perceived ability to lift mood, cause euphoria or a “high,” or to improve functioning. Two-thirds of cases involved patients with preexisting substance misuse. Placing the case reports in the context of the millions of patients prescribed antidepressants during this period, Haddad concluded the “incidence of [antidepressant] addiction [is] so low as to be clinically irrelevant.”¹⁷

Despite this conclusion, Haddad singled out amineptine and tranylcypromine as antidepressants with some evidence of true addictive potential.^17,18 A more recent case series described 14 patients who met DSM-IV criteria for substance abuse of tertiary amine TCAs (which have strong anticholinergic activity) and concluded that “misuse of [TCAs] is more common than generally appreciated.”¹⁹ In keeping with that claim, a study of 54 outpatients taking unspecified antidepressants found that up to 15% met DSM-III-R criteria for substance dependence (for the antidepressant) in the past year, although that rate was much lower than the rate of benzodiazepine dependence (47%) in a comparative sample.²⁰ Finally, a comprehensive review by Evans and Sullivan²¹ found anecdotal reports published before 2014 that detailed misuse, abuse, and dependence with MAOIs, TCAs, fluoxetine, venlafaxine, bupropion, tianeptine, and amineptine. Taken together, existing evidence indicates that select individuals—typically those with other SUD comorbidity—sometimes misuse antidepressants in a way that suggests addiction.

Still, while it is well known that abrupt cessation of antidepressants can result in a discontinuation syndrome characterized by flu-like symptoms, nausea, and dizziness,²² physiologic withdrawal effects must be distinguished from historical definitions of substance “abuse” and the broader concept of psychological “addiction” or drug dependence^18,23 now incorporated into the DSM-5 definition of SUDs.²⁴ Indeed, although withdrawal symptoms were reported by more than half of those who took antidepressants and responded to a recent online survey,²⁵ evidence to support the existence of significant antidepressant tolerance, craving, or compulsive use is lacking.^17,18 Antidepressants as a class do not appear to be significantly rewarding or reinforcing and, on the contrary, discontinuation by patients is common in clinical practice.²⁶ The popular claim that some individuals taking antidepressants “can’t quit”²⁷ must also be disentangled from loss of therapeutic effects upon cessation.

Bupropion. A more convincing argument for antidepressant addiction can be made for bupropion, a weak norepinephrine and dopamine reuptake inhibitor with an otherwise unclear mechanism of action.²⁸ In 2002, the first report of recreational bupropion M/A described a 13-year-old girl who took 2,400 mg orally (recommended maximum dose is 450 mg/d in divided doses) after being told it would give her “a better high than amphetamine.”²⁹ This was followed in the same year by the first report of recreational M/A of bupropion via nasal insufflation (snorting), resulting in a seizure,³⁰ and in 2013 by the first published case of M/A by IV self-administration.³¹

Continue to: The M/A potential of bupropion...

The M/A potential of bupropion, most commonly via intranasal administration, is now broadly recognized based on several case reports describing desired effects that include a euphoric high and a stimulating “buzz” similar to that of cocaine or methamphetamine but less intense.^29-36 Among recreational users, bupropion tablets are referred to as “welbys,” “wellies,” “dubs,” or “barnies.”³⁷ Media coverage of a 2013 outbreak of bupropion M/A in Toronto detailed administration by snorting, smoking, and injection, and described bupropion as “poor man’s cocaine.”³⁸ Between 2003 and 2016, 2,232 cases of bupropion misuse/abuse/dependence adverse drug reactions were reported to the European Monitoring Agency.³⁷ A review of intentional bupropion M/A reported to US Poison Control Centers between 2000 to 2013 found 975 such cases, with the yearly number tripling between 2000 and 2012.³⁹ In this sample, nearly half (45%) of the users were age 13 to 19, and 76% of cases involved oral ingestion. In addition to bupropion M/A among younger people, individuals who misuse bupropion often include those with existing SUDs but limited access to illicit stimulants and those trying to evade detection by urine toxicology screening.³³ For example, widespread use and diversion has been well documented within correctional settings, and as a result, many facilities have removed bupropion from their formularies.^{21,28,33,34,40}

Beyond desired effects, the most common adverse events associated with bupropion M/A are listed in Table 2,^{28,30,32-34,36,39} along with their incidence based on cases brought to the attention of US Poison Control Centers.³⁹ With relatively little evidence of a significant bupropion withdrawal syndrome,³⁷ the argument in favor of modeling bupropion as a truly addictive drug is limited to anecdotal reports of cravings and compulsive self-administration³⁵ and pro-dopaminergic activity (reuptake inhibition) that might provide a mechanism for potential rewarding and reinforcing effects.⁴⁰ While early preclinical studies of bupropion failed to provide evidence of amphetamine-like abuse potential,^41,42 non-oral administration in amounts well beyond therapeutic dosing could account for euphoric effects and a greater risk of psychological dependence and addiction.^21,28,40

Bupropion also has an FDA indication as an aid to smoking cessation treatment, and the medication demonstrated early promise in the pharmacologic treatment of psycho­stimulant use disorders, with reported improvements in cravings and other SUD outcomes.^43-45 However, subsequent randomized controlled trials (RCTs) failed to demonstrate a clear therapeutic role for bupropion in the treatment of cocaine^46,47 and methamphetamine use disorders (although some secondary analyses suggest possible therapeutic effects among non-daily stimulant users who are able to maintain good adherence with bupropion).^48-51 Given these overall discouraging results, the additive seizure risk of bupropion use with concomitant psychostimulant use, and the potential for M/A and diversion of bupropion (particularly among those with existing SUDs), the use of bupropion for the off-label treatment of stimulant use disorders is not advised.

Antipsychotics

As dopamine antagonists, antipsychotics are typically considered to have low potential for rewarding or reinforcing effects. Indeed, misuse of antipsychotics was a rarity in the first-generation era, with only a few published reports of haloperidol M/A within a small cluster of naïve young people who developed acute EPS,⁵² and a report of diversion in a prison with the “sadistic” intent of inflicting dystonic reactions on others.⁵³ A more recent report described 2additional cases of M/A involving haloperidol and trifluoperazine.⁵⁴ Some authors have described occasional drug-seeking behavior for low-potency D2 blockers such as chlorpromazine, presumably based on their M/A as anticholinergic medications.⁵⁵

The potential for antipsychotic M/A has gained wider recognition since the advent of the SGAs. Three cases of prescription olanzapine M/A have been published to date. One involved a man who malingered manic symptoms to obtain olanzapine, taking ≥40 mg at a time (beyond his prescribed dose of 20 mg twice daily) to get a “buzz,” and combining it with alcohol and benzodiazepines for additive effects or to “come down” from cocaine.⁵⁶ This patient noted that olanzapine was “a popular drug at parties” and was bought, sold, or traded among users, and occasionally administered intravenously. Two other cases described women who self-administered olanzapine, 40 to 50 mg/d, for euphoric and anxiolytic effects.^57,58 James et al⁵⁹ detailed a sample of 28 adults who reported “non-medical use” of olanzapine for anxiolytic effects, as a sleep aid, or to “escape from worries.”

Continue to: Quetiapine

Quetiapine. In contrast to some reports of olanzapine M/A in which the line between M/A and “self-medication” was blurred, quetiapine has become a more convincing example of clear recreational antipsychotic M/A. Since the first report of oral and intranasal quetiapine M/A in the Los Angeles County Jail published in 2004,⁵⁵ subsequent cases have detailed other novel methods of recreational self-administration^60-68 (Table 3^55,60-68), and additional reports have been published in non-English language journals.^69,70 Collectively, these case reports have detailed that quetiapine is:

• misused for primary subjective effects as well as to mitigate the unpleasant effects of other drugs^60,67
• referred to as “quell,”“Q,” “Susie-Q,” “squirrel,” and “baby heroin”^55,71,72
• often obtained by malingering psychiatric symptoms^55,61,63,65
• diverted/sold with “street value” both within and outside of psychiatric facilities and correctional settings.^{55,60-62,67,68,73}

These anecdotal accounts of quetiapine M/A have since been corroborated on a larger scale based on several retrospective studies. Although early reports of quetiapine M/A occurring in correctional settings have resulted in formulary removal,^71,74 quetiapine M/A is by no means limited to forensic populations and is especially common among those with comorbid SUDs. A survey of 74 patients enrolled in a Canadian methadone program reported that nearly 60% had misused quetiapine at some point.⁷⁵ Among an Australian sample of 868 individuals with active IV drug abuse, 31% reported having misused quetiapine.⁷⁶ Finally, within a small sample of patients with SUDs admitted to a detoxification unit in New York City, 17% reported M/A of SGAs.⁷⁷ In this study, SGAs were often taken in conjunction with other drugs of abuse in order to “recover” from or “enhance” the effects of other drugs or to “experiment.” Quetiapine was by far the most frequently abused SGA, reported in 96% of the sample; the most frequently reported SGA/drug combinations were quetiapine/alcohol/opioids, quetiapine/cocaine, and quetiapine/opioids.

Looking more broadly at poison center data, reports to the US National Poison Data System (NPDS) from 2005 to 2011 included 3,116 cases of quetiapine abuse (37.5%, defined as intentional recreational use in order to obtain a “high”) or misuse (62.5%, defined as improper use or dosing for non-recreational purposes).⁷⁸ A more recent analysis of NPDS reports from 2003 to 2013 found 2,118 cases of quetiapine abuse, representing 61% of all cases of reported SGA abuse.⁷⁹ An analysis of the European Medicines Agency Adverse Drug Database yielded 18,112 reports of quetia­pine misuse, abuse, dependence, and withdrawal for quetiapine (from 2005 to 2016) compared with 4,178 for olanzapine (from 2004 to 2016).⁸⁰ These reports identified 368 fatalities associated with quetiapine.

The rate of quetiapine M/A appears to be increasing sharply. Reports of quetiapine M/A to poison centers in Australia increased nearly 7-fold from 2006 to 2016.⁸¹ Based on reports to the Drug Abuse Warning System, US emergency department visits for M/A of quetiapine increased from 19,195 in 2005 to 32,024 in 2011 (an average of 27,114 visits/year), with 75% of cases involving quetiapine taken in combination with other prescription drugs, alcohol, or illicit drugs.⁸² Consistent with poison center data, M/A was reported for other antipsychotics, but none nearly as frequently as for quetiapine.

With increasingly frequent quetiapine M/A, clinicians should be vigilant in monitoring for medical morbidity related to quetiapine and cumulative toxicity with other drugs. The most frequent adverse events associated with quetiapine M/A reported to US Poison Control Centers are presented in Table 4.^78,79

Continue to: Unlike bupropion...

Unlike bupropion, quetiapine’s dopamine antagonism makes it unlikely to be a truly addictive drug, although this mechanism of action could mediate an increase in concurrent psychostimulant use.⁸³ A few case reports have described a quetiapine discontinuation syndrome similar to that of antidepressants,^60,65,84-88 but withdrawal symptoms suggestive of physiologic dependence may be mediated by non-dopaminergic effects through histamine and serotonin receptors.^84,89 Evidence for quetiapine misuse being associated with craving and compulsive use is lacking, and true quetiapine addiction is probably rare.

Similar to bupropion, preliminary findings have suggested promise for quetiapine as a putative therapy for other SUDs.^90-93 However, subsequent RCTs have failed to demonstrate a therapeutic effect for alcohol and cocaine use disorders.^94-96 Given these negative results and the clear M/A potential of quetiapine, off-label use of quetiapine for the treatment of SUDs and psychiatric symptoms among those with SUDs must be considered judiciously, with an eye towards possible diversion and avoiding the substitution of one drug of abuse for another.

Gabapentinoids

In 1997, the first published case report of gabapentin M/A described a woman who self-administered her husband’s gabapentin to reduce cravings for and withdrawal from cocaine.⁹⁷ The authors highlighted the possible therapeutic benefit of gabapentin in this regard rather than raising concerns about diversion and M/A. By 2004, however, reports of recreational gabapentin M/A emerged among inmates incarcerated within Florida correctional facilities who self-administered intranasal gabapentin to achieve a “high” that was “reminiscent of prior effects from intranasal ingestion of cocaine powder.”⁹⁸ In 2007, a single case of gabapentin misuse up to 7,200 mg/d (recommended dosing is ≤3,600 mg/d) was reported, with documentation of both tolerance and withdrawal symptoms.⁹⁹ As of 2017, a total of 36 cases of gabapentin M/A and 19 cases of pregabalin M/A have been published.¹⁰⁰

In the past decade, anecdotal reports have given way to larger-scale epidemiologic data painting a clear picture of the now-widespread M/A of gabapentin and other gabapentinoids. For example, a study of online descriptions of gabapentin and pregabalin M/A from 2008 to 2010 documented:

• oral and IM use (gabapentin)
• IV and rectal (“plugging”) use (pregabalin)
• “parachuting” (emptying the contents of capsules for a larger dose) (pregabalin)
• euphoric, entactogenic, stimulant, calming/anxiolytic, and dissociative subjective effects (gabapentin/pregabalin)
• rapid development of tolerance to euphoric effects leading to self-administration of increasing doses (gabapentin/pregabalin)
• frequent co-administration with other drugs of abuse, including alcohol, benzodiazepines, cannabis, stimulants, opiates, hallucinogens, gamma-hydroxybutyrate, mephedrone, and Salvia divinorum (gabapentin/pregabalin)¹⁰¹

Several systematic reviews of both anecdotal reports and epidemiologic studies published in the past few years provide additional evidence of the above, such as:

• excessive dosing with self-administration
• intranasal and inhaled routes of administration
• diversion and “street value”
• greater M/A potential of pregabalin than gabapentin
• the presence of gabapentinoids in postmortem toxicology analyses, suggesting a role in overdose fatalities when combined with other drugs.^100,102,103

Continue to: The European Medicine Agency's EudraVigilance database...

The European Medicine Agency’s EudraVigilance database included 4,301 reports of gabapentin misuse, abuse, or dependence, and 7,639 such reports for pregabalin, from 2006 to 2015 (rising sharply after 2012), with 86 gabapentin-related and 27 pregabalin-related fatalities.¹⁰⁴ Data from the Drug Diversion Program of the Researched Abuse, Diversion, and Addiction-Related Surveillance System from 2002 to 2015 have likewise revealed that gabapentin diversion increased significantly in 2013.¹⁰⁵

While the prevalence of gabapentinoid M/A is not known, rates appear to be significantly lower than for traditional drugs of abuse such as cannabis, cocaine, 3,4-methylenedioxymethamphetamine (MDMA), and opioids.^106,107 However, gabapentin and pregabalin M/A appears to be increasingly common among individuals with SUDs and in particular among those with opioid use disorders (OUDs). For example, a 2015 report indicated that 15% of an adult cohort in Appalachian Kentucky with nonmedical use of diverted prescription opioids reported gabapentin M/A, an increase of nearly 3,000% since 2008.¹⁰⁸ Based on data from a US insurance enrollment and claims database, researchers found that the rate of gabapentin overuse among those also overusing opioids was 12% compared with only 2% for those using gabapentin alone.¹⁰⁹ It has also been reported that gabapentin is sometimes used as a “cutting agent” for heroin.¹¹⁰

Those who use gabapentinoids together with opioids report that gabapentin and pregabalin potentiate the euphoric effects of methadone¹¹¹ and endorse specific beliefs that pregabalin increases both the desired effects of heroin as well as negative effects such as “blackouts,” loss of control, and risk of overdose.¹¹² Indeed, sustained M/A of gabapentin and opioids together has been found to increase emergency department utilization, drug-related hospitalization, and respiratory depression.¹¹³ Based on a case-control study of opioid users in Canada, co-prescription of gabapentin and opioids was associated with a 50% increase in death from opioid-related causes compared with prescription of opioids alone.¹¹⁴

Case reports documenting tolerance, withdrawal, craving, and loss of control suggest a true addictive potential for gabapentinoids, but Bonnet and Sherbaum¹⁰⁰ concluded that while there is robust evidence of abusers “liking” gabapentin and pregabalin (eg, reward), evidence of “wanting” them (eg, psychological dependence) in the absence of other SUDs has been limited to only a few anecdotal reports with pregabalin. Accordingly, the risk of true addiction to gabapentinoids by those without preexisting SUDs appears to be low. Nonetheless, the M/A potential of both gabapentin and pregabalin is clear and in the context of a nationwide opioid epidemic, the increased morbidity/mortality risk related to combined use of gabapentinoids and opioids is both striking and concerning. Consequently, the state of Kentucky recently recognized the M/A potential of gabapentin by designating it a Schedule V controlled substance (pregabalin is already a Schedule V drug according to the US Drug Enforcement Agency),^103,113 and several other states now mandate the reporting of gabapentin prescriptions to prescription drug monitoring programs.¹¹⁵

Following a similar pattern to antidepressants and antipsychotics, a potential role for gabapentin in the treatment of cocaine use disorders was supported in preliminary studies,^116-118 but not in subsequent RCTs.^119-121 However, there is evidence from RCTs to support the use of gabapentin and pregabalin in the treatment of alcohol use disorders.^122-124 Gabapentin was also found to significantly reduce cannabis use and withdrawal symptoms in patients compared with placebo in an RCT of individuals with cannabis use disorders.¹²⁵ The perceived safety of gabapentinoids by clinicians, their subjective desirability by patients with SUDs, and efficacy data supporting a therapeutic role in SUDs must be balanced with recognition that approximately 80% of gabapentin prescriptions are written for off-label indications for which there is little supporting evidence,¹⁰⁹ such as low back pain.¹²⁶ Clinicians considering prescribing gabapentinoids to manage psychiatric symptoms, such as anxiety and insomnia, should carefully consider the risk of M/A and other potential morbidities, especially in the setting of SUDs and OUD in particular.

Continue to: Problematic, even if not addictive

Problematic, even if not addictive

It is sometimes claimed that “addiction” to psychiatric medications is not limited to stimulants and benzodiazepines.^27,127 Although anticholinergics, antidepressants, antipsychotics, and gabapentinoids can be drugs of abuse, with some users reporting physiologic withdrawal upon discontinuation, there is only limited evidence that the M/A of these psychiatric medications is associated with the characteristic features of a more complete definition of “addiction,” which may include:

• inability to consistently abstain
• impairment in behavioral control
• diminished recognition of significant problems associated with use
• a dysfunctional emotional response to chronic use.¹²⁸

Nonetheless, the literature documenting anticholinergic, antidepressant, antipsychotic, and gabapentinoid M/A includes several common features, including:

• initial reports among those with limited access to illicit drugs (eg, young people and incarcerated individuals) and subsequent spread to a wider population with more unconventional routes of administration
• use for recreational purposes and other subjective pseudo-therapeutic effects, often in combination with alcohol and illicit drugs
• greater M/A potential of certain medications within each of these drug classes (eg, trihexyphenidyl, bupropion, quetiapine)
• malingering psychiatric symptoms in order to obtain medications from prescribers and diversion for black market sale
• observations that medications might constitute therapy for SUDs that were not supported in subsequent RCTs (with the exception of gabapentin for alcohol and cannabis use disorders)
• increasing evidence of toxicity related to M/A, which suggests that prescription by clinicians has limited benefit and high risk for patients with SUDs.

Bottom Line

Some psychiatric medications are taken as drugs of abuse. Clinicians should be particularly aware of the misuse/abuse potential of anticholinergics, antidepressants, antipsychotics, and gabapentinoids, and use them cautiously, if at all, when treating patients with existing substance use disorders.

Related Resources

• Substance Abuse and Mental Health Services Administration. Prescription drug misuse and abuse. https://www.samhsa.gov/topics/prescription-drug-misuse-abuse.
• Substance Abuse and Mental Health Services Administration. Types of commonly misused or abused drugs. https://www.samhsa.gov/prescription-drug-misuse-abuse/types.
• National Institute on Drug Abuse. Misuse of prescription drugs. https://www.drugabuse.gov/publications/research-reports/misuse-prescription-drugs/summary.
• National Institute on Drug Abuse. New clinician screening tool available for substance use. https://www.drugabuse.gov/news-events/news-releases/2018/06/newclinician-screening-tool-available-substance-use.

Drug Brand Names

Amitriptyline • Elavil, Endep
Benztropine • Cogentin
Biperiden • Akineton
Bupropion • Wellbutrin, Zyban
Chlorpromazine • Thorzine
Fluoxetine • Prozac
Haloperidol • Haldol
Olanzapine • Zyprexa
Orphenadrine • Disipal, Norflex
Pregabalin • Lyrica, Lyrica CR
Procyclidine • Kemadrin
Quetiapine • Seroquel
Tianeptine • Coaxil, Stablon
Tranylcypromine • Parnate
Trifluoperazine • Stelazine
Trihexyphenidyl • Artane, Tremin
Venlafaxine • Effexor

While some classes of medications used to treat psychi­atric disorders, such as stimulants and benzodiazepines, are well-recognized as controlled substances and drugs of abuse, clinicians may be less familiar with the potential misuse/abuse of other psychiatric medications. This article reviews the evidence related to the misuse/abuse of anticholinergics, antidepressants, antipsychotics, and gabapentinoids.

The terms “misuse,” “abuse,” and “addiction” are used variably in the literature without standardized definitions. For this review, “misuse/abuse (M/A)” will be used to collectively describe self-administration that is recreational or otherwise inconsistent with legal or medical guidelines, unless a specific distinction is made. Whether or not the medications reviewed are truly “addictive” will be briefly discussed for each drug class, but the focus will be on clinically relevant aspects of M/A, including:

• excessive self-administration
• self-administration by non-oral routes
• co-administration with other drugs of abuse
• malingering of psychiatric symptoms to obtain prescriptions
• diversion for sale to third parties
• toxicity from overdose.

Anticholinergic medications

The first case describing the deliberate M/A of an anticholinergic medication for its euphoric effects was published in 1960.Further reportsfollowed in Europe before the M/A potential of prescription anticholinergic medications among psychiatric patients with an overdose syndrome characterized by atropinism and toxic psychosis was more widely recognized in the United States in the 1970s. Most reported cases of M/A to date have occurred among patients with psychiatric illness because anticholinergic medications, including trihexyphenidyl, benztropine, biperiden, procyclidine, and orphenadrine, were commonly prescribed for the management of first-generation and high dopamine D2-affinity antipsychotic-induced extrapyramidal symptoms (EPS). For example, one study of 234 consecutively hospitalized patients with schizophrenia noted an anticholinergic M/A incidence of 6.5%.¹

However, anticholinergic M/A is not limited to individuals with psychotic disorders. A UK study of 154 admissions to an inpatient unit specializing in behavioral disturbances found a 12-month trihexyphenidyl M/A incidence of 17%; the most common diagnosis among abusers was antisocial personality disorder.² Anticholinergic M/A has also been reported among patients with a primary diagnosis of substance use disorders (SUDs)³ as well as more indiscriminately in prison settings,⁴ with some inmates exchanging trihexyphenidyl as currency and using it recreationally by crushing it into powder and smoking it with tobacco.⁵ Others have noted that abusers sometimes take anticholinergics with alcohol in order to “potentiate” the effects of each substance.^6,7 Pullen et al⁸ described individuals with and without psychiatric illness who stole anticholinergic medications, purchased them from other patients, or bought them “on the street.” Malingering EPS in order to obtain anticholinergic medications has also been well documented.⁹ Clearly, anticholinergic M/A can occur in psychiatric and non-psychiatric populations, both within and outside of clinical settings. Although anticholinergic M/A appears to be less frequent in the United States now that second-generation antipsychotics (SGAs) are more frequently prescribed, M/A remains common in some settings outside of the United States.⁷

Among the various anticholinergic medications prescribed for EPS, trihexyphenidyl has been reported to have the greatest M/A potential, which has been attributed to its potency,¹⁰ its stimulating effects (whereas benztropine is more sedating),¹¹ and its former popularity among prescribers.⁸ Marken et al¹¹ published a review of 110 reports of M/A occurring in patients receiving anticholinergic medications as part of psychiatric treatment in which 69% of cases involved taking trihexyphenidyl 15 to 60 mg at a time (recommended dosing is 6 to 10 mg/d in divided doses).Most of these patients were prescribed anticholinergic medications for diagnostically appropriate reasons—only 7% were described as “true abusers” with no medical indication. Anticholinergic M/A was typically driven by a desire for euphoric and psychedelic/hallucinogenic effects, although in some cases, anticholinergic M/A was attributed to self-medication of EPS and depressive symptoms. These findings illustrate the blurred distinction between recreational use and perceived subjective benefit, and match those of a subsequent study of 50 psychiatric patients who reported anticholinergic M/A not only to “get high,” but to “decrease depression,” “increase energy,” and decrease antipsychotic adverse effects.¹² Once again, trihexyphenidyl was the most frequently misused anticholinergic in this sample.

Table 1^{2,3,7,8,10-15} outlines the subjective effects sought and experienced by anticholinergic abusers as well as potential toxic effects; there is the potential for overlap. Several authors have also described physiologic dependence with long-term trihexyphenidyl use, including tolerance and a withdrawal/abstinence syndrome.^7,16 In addition, there have been several reports of coma¹³ and death in the setting of intended suicide by overdose of anticholinergic medications.^14,15

Although anticholinergic M/A in the United States now appears to be less common, clinicians should remain aware of the M/A potential of anticholinergic medications prescribed for EPS. Management of M/A involves:

• detection
• reducing anticholinergic exposure by managing EPS with alternative strategies, such as switching or reducing the dose of the antipsychotic medication
• gradual tapering of anticholinergic medications to minimize withdrawal.¹¹

Continue to: Antidepressants

Haddad¹⁷ published a review of 21 English-language case reports from 1966 to 1998 describing antidepressant use in which individuals met DSM-IV criteria for substance dependence to the medication. An additional 14 cases of antidepressant M/A were excluded based on insufficient details to support a diagnosis of dependence. The 21 reported cases involved:

• tranylcypromine (a monoamine oxidase inhibitor [MAOI])
• amitriptyline (a tricyclic antidepressant [TCA])
• fluoxetine (a selective serotonin reuptake inhibitor [SSRI])
• amineptine (a TCA previously available in France but removed from the market in 1999 in part due to its abuse potential)
• nomifensine (a norepinephrine/dopamine reuptake inhibitor previously available in the United Kingdom but removed in 1986 due to hemolytic anemia).

In 95% of cases, the antidepressants were prescribed for treatment of an affective disorder but were abused for stimulant effects or the perceived ability to lift mood, cause euphoria or a “high,” or to improve functioning. Two-thirds of cases involved patients with preexisting substance misuse. Placing the case reports in the context of the millions of patients prescribed antidepressants during this period, Haddad concluded the “incidence of [antidepressant] addiction [is] so low as to be clinically irrelevant.”¹⁷

Despite this conclusion, Haddad singled out amineptine and tranylcypromine as antidepressants with some evidence of true addictive potential.^17,18 A more recent case series described 14 patients who met DSM-IV criteria for substance abuse of tertiary amine TCAs (which have strong anticholinergic activity) and concluded that “misuse of [TCAs] is more common than generally appreciated.”¹⁹ In keeping with that claim, a study of 54 outpatients taking unspecified antidepressants found that up to 15% met DSM-III-R criteria for substance dependence (for the antidepressant) in the past year, although that rate was much lower than the rate of benzodiazepine dependence (47%) in a comparative sample.²⁰ Finally, a comprehensive review by Evans and Sullivan²¹ found anecdotal reports published before 2014 that detailed misuse, abuse, and dependence with MAOIs, TCAs, fluoxetine, venlafaxine, bupropion, tianeptine, and amineptine. Taken together, existing evidence indicates that select individuals—typically those with other SUD comorbidity—sometimes misuse antidepressants in a way that suggests addiction.

Still, while it is well known that abrupt cessation of antidepressants can result in a discontinuation syndrome characterized by flu-like symptoms, nausea, and dizziness,²² physiologic withdrawal effects must be distinguished from historical definitions of substance “abuse” and the broader concept of psychological “addiction” or drug dependence^18,23 now incorporated into the DSM-5 definition of SUDs.²⁴ Indeed, although withdrawal symptoms were reported by more than half of those who took antidepressants and responded to a recent online survey,²⁵ evidence to support the existence of significant antidepressant tolerance, craving, or compulsive use is lacking.^17,18 Antidepressants as a class do not appear to be significantly rewarding or reinforcing and, on the contrary, discontinuation by patients is common in clinical practice.²⁶ The popular claim that some individuals taking antidepressants “can’t quit”²⁷ must also be disentangled from loss of therapeutic effects upon cessation.

Bupropion. A more convincing argument for antidepressant addiction can be made for bupropion, a weak norepinephrine and dopamine reuptake inhibitor with an otherwise unclear mechanism of action.²⁸ In 2002, the first report of recreational bupropion M/A described a 13-year-old girl who took 2,400 mg orally (recommended maximum dose is 450 mg/d in divided doses) after being told it would give her “a better high than amphetamine.”²⁹ This was followed in the same year by the first report of recreational M/A of bupropion via nasal insufflation (snorting), resulting in a seizure,³⁰ and in 2013 by the first published case of M/A by IV self-administration.³¹

Continue to: The M/A potential of bupropion...

The M/A potential of bupropion, most commonly via intranasal administration, is now broadly recognized based on several case reports describing desired effects that include a euphoric high and a stimulating “buzz” similar to that of cocaine or methamphetamine but less intense.^29-36 Among recreational users, bupropion tablets are referred to as “welbys,” “wellies,” “dubs,” or “barnies.”³⁷ Media coverage of a 2013 outbreak of bupropion M/A in Toronto detailed administration by snorting, smoking, and injection, and described bupropion as “poor man’s cocaine.”³⁸ Between 2003 and 2016, 2,232 cases of bupropion misuse/abuse/dependence adverse drug reactions were reported to the European Monitoring Agency.³⁷ A review of intentional bupropion M/A reported to US Poison Control Centers between 2000 to 2013 found 975 such cases, with the yearly number tripling between 2000 and 2012.³⁹ In this sample, nearly half (45%) of the users were age 13 to 19, and 76% of cases involved oral ingestion. In addition to bupropion M/A among younger people, individuals who misuse bupropion often include those with existing SUDs but limited access to illicit stimulants and those trying to evade detection by urine toxicology screening.³³ For example, widespread use and diversion has been well documented within correctional settings, and as a result, many facilities have removed bupropion from their formularies.^{21,28,33,34,40}

Beyond desired effects, the most common adverse events associated with bupropion M/A are listed in Table 2,^{28,30,32-34,36,39} along with their incidence based on cases brought to the attention of US Poison Control Centers.³⁹ With relatively little evidence of a significant bupropion withdrawal syndrome,³⁷ the argument in favor of modeling bupropion as a truly addictive drug is limited to anecdotal reports of cravings and compulsive self-administration³⁵ and pro-dopaminergic activity (reuptake inhibition) that might provide a mechanism for potential rewarding and reinforcing effects.⁴⁰ While early preclinical studies of bupropion failed to provide evidence of amphetamine-like abuse potential,^41,42 non-oral administration in amounts well beyond therapeutic dosing could account for euphoric effects and a greater risk of psychological dependence and addiction.^21,28,40

Bupropion also has an FDA indication as an aid to smoking cessation treatment, and the medication demonstrated early promise in the pharmacologic treatment of psycho­stimulant use disorders, with reported improvements in cravings and other SUD outcomes.^43-45 However, subsequent randomized controlled trials (RCTs) failed to demonstrate a clear therapeutic role for bupropion in the treatment of cocaine^46,47 and methamphetamine use disorders (although some secondary analyses suggest possible therapeutic effects among non-daily stimulant users who are able to maintain good adherence with bupropion).^48-51 Given these overall discouraging results, the additive seizure risk of bupropion use with concomitant psychostimulant use, and the potential for M/A and diversion of bupropion (particularly among those with existing SUDs), the use of bupropion for the off-label treatment of stimulant use disorders is not advised.

Antipsychotics

As dopamine antagonists, antipsychotics are typically considered to have low potential for rewarding or reinforcing effects. Indeed, misuse of antipsychotics was a rarity in the first-generation era, with only a few published reports of haloperidol M/A within a small cluster of naïve young people who developed acute EPS,⁵² and a report of diversion in a prison with the “sadistic” intent of inflicting dystonic reactions on others.⁵³ A more recent report described 2additional cases of M/A involving haloperidol and trifluoperazine.⁵⁴ Some authors have described occasional drug-seeking behavior for low-potency D2 blockers such as chlorpromazine, presumably based on their M/A as anticholinergic medications.⁵⁵

The potential for antipsychotic M/A has gained wider recognition since the advent of the SGAs. Three cases of prescription olanzapine M/A have been published to date. One involved a man who malingered manic symptoms to obtain olanzapine, taking ≥40 mg at a time (beyond his prescribed dose of 20 mg twice daily) to get a “buzz,” and combining it with alcohol and benzodiazepines for additive effects or to “come down” from cocaine.⁵⁶ This patient noted that olanzapine was “a popular drug at parties” and was bought, sold, or traded among users, and occasionally administered intravenously. Two other cases described women who self-administered olanzapine, 40 to 50 mg/d, for euphoric and anxiolytic effects.^57,58 James et al⁵⁹ detailed a sample of 28 adults who reported “non-medical use” of olanzapine for anxiolytic effects, as a sleep aid, or to “escape from worries.”

Continue to: Quetiapine

Quetiapine. In contrast to some reports of olanzapine M/A in which the line between M/A and “self-medication” was blurred, quetiapine has become a more convincing example of clear recreational antipsychotic M/A. Since the first report of oral and intranasal quetiapine M/A in the Los Angeles County Jail published in 2004,⁵⁵ subsequent cases have detailed other novel methods of recreational self-administration^60-68 (Table 3^55,60-68), and additional reports have been published in non-English language journals.^69,70 Collectively, these case reports have detailed that quetiapine is:

• misused for primary subjective effects as well as to mitigate the unpleasant effects of other drugs^60,67
• referred to as “quell,”“Q,” “Susie-Q,” “squirrel,” and “baby heroin”^55,71,72
• often obtained by malingering psychiatric symptoms^55,61,63,65
• diverted/sold with “street value” both within and outside of psychiatric facilities and correctional settings.^{55,60-62,67,68,73}

These anecdotal accounts of quetiapine M/A have since been corroborated on a larger scale based on several retrospective studies. Although early reports of quetiapine M/A occurring in correctional settings have resulted in formulary removal,^71,74 quetiapine M/A is by no means limited to forensic populations and is especially common among those with comorbid SUDs. A survey of 74 patients enrolled in a Canadian methadone program reported that nearly 60% had misused quetiapine at some point.⁷⁵ Among an Australian sample of 868 individuals with active IV drug abuse, 31% reported having misused quetiapine.⁷⁶ Finally, within a small sample of patients with SUDs admitted to a detoxification unit in New York City, 17% reported M/A of SGAs.⁷⁷ In this study, SGAs were often taken in conjunction with other drugs of abuse in order to “recover” from or “enhance” the effects of other drugs or to “experiment.” Quetiapine was by far the most frequently abused SGA, reported in 96% of the sample; the most frequently reported SGA/drug combinations were quetiapine/alcohol/opioids, quetiapine/cocaine, and quetiapine/opioids.

Looking more broadly at poison center data, reports to the US National Poison Data System (NPDS) from 2005 to 2011 included 3,116 cases of quetiapine abuse (37.5%, defined as intentional recreational use in order to obtain a “high”) or misuse (62.5%, defined as improper use or dosing for non-recreational purposes).⁷⁸ A more recent analysis of NPDS reports from 2003 to 2013 found 2,118 cases of quetiapine abuse, representing 61% of all cases of reported SGA abuse.⁷⁹ An analysis of the European Medicines Agency Adverse Drug Database yielded 18,112 reports of quetia­pine misuse, abuse, dependence, and withdrawal for quetiapine (from 2005 to 2016) compared with 4,178 for olanzapine (from 2004 to 2016).⁸⁰ These reports identified 368 fatalities associated with quetiapine.

The rate of quetiapine M/A appears to be increasing sharply. Reports of quetiapine M/A to poison centers in Australia increased nearly 7-fold from 2006 to 2016.⁸¹ Based on reports to the Drug Abuse Warning System, US emergency department visits for M/A of quetiapine increased from 19,195 in 2005 to 32,024 in 2011 (an average of 27,114 visits/year), with 75% of cases involving quetiapine taken in combination with other prescription drugs, alcohol, or illicit drugs.⁸² Consistent with poison center data, M/A was reported for other antipsychotics, but none nearly as frequently as for quetiapine.

With increasingly frequent quetiapine M/A, clinicians should be vigilant in monitoring for medical morbidity related to quetiapine and cumulative toxicity with other drugs. The most frequent adverse events associated with quetiapine M/A reported to US Poison Control Centers are presented in Table 4.^78,79

Continue to: Unlike bupropion...

Unlike bupropion, quetiapine’s dopamine antagonism makes it unlikely to be a truly addictive drug, although this mechanism of action could mediate an increase in concurrent psychostimulant use.⁸³ A few case reports have described a quetiapine discontinuation syndrome similar to that of antidepressants,^60,65,84-88 but withdrawal symptoms suggestive of physiologic dependence may be mediated by non-dopaminergic effects through histamine and serotonin receptors.^84,89 Evidence for quetiapine misuse being associated with craving and compulsive use is lacking, and true quetiapine addiction is probably rare.

Similar to bupropion, preliminary findings have suggested promise for quetiapine as a putative therapy for other SUDs.^90-93 However, subsequent RCTs have failed to demonstrate a therapeutic effect for alcohol and cocaine use disorders.^94-96 Given these negative results and the clear M/A potential of quetiapine, off-label use of quetiapine for the treatment of SUDs and psychiatric symptoms among those with SUDs must be considered judiciously, with an eye towards possible diversion and avoiding the substitution of one drug of abuse for another.

Gabapentinoids

In 1997, the first published case report of gabapentin M/A described a woman who self-administered her husband’s gabapentin to reduce cravings for and withdrawal from cocaine.⁹⁷ The authors highlighted the possible therapeutic benefit of gabapentin in this regard rather than raising concerns about diversion and M/A. By 2004, however, reports of recreational gabapentin M/A emerged among inmates incarcerated within Florida correctional facilities who self-administered intranasal gabapentin to achieve a “high” that was “reminiscent of prior effects from intranasal ingestion of cocaine powder.”⁹⁸ In 2007, a single case of gabapentin misuse up to 7,200 mg/d (recommended dosing is ≤3,600 mg/d) was reported, with documentation of both tolerance and withdrawal symptoms.⁹⁹ As of 2017, a total of 36 cases of gabapentin M/A and 19 cases of pregabalin M/A have been published.¹⁰⁰

In the past decade, anecdotal reports have given way to larger-scale epidemiologic data painting a clear picture of the now-widespread M/A of gabapentin and other gabapentinoids. For example, a study of online descriptions of gabapentin and pregabalin M/A from 2008 to 2010 documented:

• oral and IM use (gabapentin)
• IV and rectal (“plugging”) use (pregabalin)
• “parachuting” (emptying the contents of capsules for a larger dose) (pregabalin)
• euphoric, entactogenic, stimulant, calming/anxiolytic, and dissociative subjective effects (gabapentin/pregabalin)
• rapid development of tolerance to euphoric effects leading to self-administration of increasing doses (gabapentin/pregabalin)
• frequent co-administration with other drugs of abuse, including alcohol, benzodiazepines, cannabis, stimulants, opiates, hallucinogens, gamma-hydroxybutyrate, mephedrone, and Salvia divinorum (gabapentin/pregabalin)¹⁰¹

Several systematic reviews of both anecdotal reports and epidemiologic studies published in the past few years provide additional evidence of the above, such as:

• excessive dosing with self-administration
• intranasal and inhaled routes of administration
• diversion and “street value”
• greater M/A potential of pregabalin than gabapentin
• the presence of gabapentinoids in postmortem toxicology analyses, suggesting a role in overdose fatalities when combined with other drugs.^100,102,103

Continue to: The European Medicine Agency's EudraVigilance database...

The European Medicine Agency’s EudraVigilance database included 4,301 reports of gabapentin misuse, abuse, or dependence, and 7,639 such reports for pregabalin, from 2006 to 2015 (rising sharply after 2012), with 86 gabapentin-related and 27 pregabalin-related fatalities.¹⁰⁴ Data from the Drug Diversion Program of the Researched Abuse, Diversion, and Addiction-Related Surveillance System from 2002 to 2015 have likewise revealed that gabapentin diversion increased significantly in 2013.¹⁰⁵

While the prevalence of gabapentinoid M/A is not known, rates appear to be significantly lower than for traditional drugs of abuse such as cannabis, cocaine, 3,4-methylenedioxymethamphetamine (MDMA), and opioids.^106,107 However, gabapentin and pregabalin M/A appears to be increasingly common among individuals with SUDs and in particular among those with opioid use disorders (OUDs). For example, a 2015 report indicated that 15% of an adult cohort in Appalachian Kentucky with nonmedical use of diverted prescription opioids reported gabapentin M/A, an increase of nearly 3,000% since 2008.¹⁰⁸ Based on data from a US insurance enrollment and claims database, researchers found that the rate of gabapentin overuse among those also overusing opioids was 12% compared with only 2% for those using gabapentin alone.¹⁰⁹ It has also been reported that gabapentin is sometimes used as a “cutting agent” for heroin.¹¹⁰

Those who use gabapentinoids together with opioids report that gabapentin and pregabalin potentiate the euphoric effects of methadone¹¹¹ and endorse specific beliefs that pregabalin increases both the desired effects of heroin as well as negative effects such as “blackouts,” loss of control, and risk of overdose.¹¹² Indeed, sustained M/A of gabapentin and opioids together has been found to increase emergency department utilization, drug-related hospitalization, and respiratory depression.¹¹³ Based on a case-control study of opioid users in Canada, co-prescription of gabapentin and opioids was associated with a 50% increase in death from opioid-related causes compared with prescription of opioids alone.¹¹⁴

Case reports documenting tolerance, withdrawal, craving, and loss of control suggest a true addictive potential for gabapentinoids, but Bonnet and Sherbaum¹⁰⁰ concluded that while there is robust evidence of abusers “liking” gabapentin and pregabalin (eg, reward), evidence of “wanting” them (eg, psychological dependence) in the absence of other SUDs has been limited to only a few anecdotal reports with pregabalin. Accordingly, the risk of true addiction to gabapentinoids by those without preexisting SUDs appears to be low. Nonetheless, the M/A potential of both gabapentin and pregabalin is clear and in the context of a nationwide opioid epidemic, the increased morbidity/mortality risk related to combined use of gabapentinoids and opioids is both striking and concerning. Consequently, the state of Kentucky recently recognized the M/A potential of gabapentin by designating it a Schedule V controlled substance (pregabalin is already a Schedule V drug according to the US Drug Enforcement Agency),^103,113 and several other states now mandate the reporting of gabapentin prescriptions to prescription drug monitoring programs.¹¹⁵

Following a similar pattern to antidepressants and antipsychotics, a potential role for gabapentin in the treatment of cocaine use disorders was supported in preliminary studies,^116-118 but not in subsequent RCTs.^119-121 However, there is evidence from RCTs to support the use of gabapentin and pregabalin in the treatment of alcohol use disorders.^122-124 Gabapentin was also found to significantly reduce cannabis use and withdrawal symptoms in patients compared with placebo in an RCT of individuals with cannabis use disorders.¹²⁵ The perceived safety of gabapentinoids by clinicians, their subjective desirability by patients with SUDs, and efficacy data supporting a therapeutic role in SUDs must be balanced with recognition that approximately 80% of gabapentin prescriptions are written for off-label indications for which there is little supporting evidence,¹⁰⁹ such as low back pain.¹²⁶ Clinicians considering prescribing gabapentinoids to manage psychiatric symptoms, such as anxiety and insomnia, should carefully consider the risk of M/A and other potential morbidities, especially in the setting of SUDs and OUD in particular.

Continue to: Problematic, even if not addictive

Problematic, even if not addictive

It is sometimes claimed that “addiction” to psychiatric medications is not limited to stimulants and benzodiazepines.^27,127 Although anticholinergics, antidepressants, antipsychotics, and gabapentinoids can be drugs of abuse, with some users reporting physiologic withdrawal upon discontinuation, there is only limited evidence that the M/A of these psychiatric medications is associated with the characteristic features of a more complete definition of “addiction,” which may include:

• inability to consistently abstain
• impairment in behavioral control
• diminished recognition of significant problems associated with use
• a dysfunctional emotional response to chronic use.¹²⁸

Nonetheless, the literature documenting anticholinergic, antidepressant, antipsychotic, and gabapentinoid M/A includes several common features, including:

• initial reports among those with limited access to illicit drugs (eg, young people and incarcerated individuals) and subsequent spread to a wider population with more unconventional routes of administration
• use for recreational purposes and other subjective pseudo-therapeutic effects, often in combination with alcohol and illicit drugs
• greater M/A potential of certain medications within each of these drug classes (eg, trihexyphenidyl, bupropion, quetiapine)
• malingering psychiatric symptoms in order to obtain medications from prescribers and diversion for black market sale
• observations that medications might constitute therapy for SUDs that were not supported in subsequent RCTs (with the exception of gabapentin for alcohol and cannabis use disorders)
• increasing evidence of toxicity related to M/A, which suggests that prescription by clinicians has limited benefit and high risk for patients with SUDs.

Bottom Line

Some psychiatric medications are taken as drugs of abuse. Clinicians should be particularly aware of the misuse/abuse potential of anticholinergics, antidepressants, antipsychotics, and gabapentinoids, and use them cautiously, if at all, when treating patients with existing substance use disorders.

Related Resources

• Substance Abuse and Mental Health Services Administration. Prescription drug misuse and abuse. https://www.samhsa.gov/topics/prescription-drug-misuse-abuse.
• Substance Abuse and Mental Health Services Administration. Types of commonly misused or abused drugs. https://www.samhsa.gov/prescription-drug-misuse-abuse/types.
• National Institute on Drug Abuse. Misuse of prescription drugs. https://www.drugabuse.gov/publications/research-reports/misuse-prescription-drugs/summary.
• National Institute on Drug Abuse. New clinician screening tool available for substance use. https://www.drugabuse.gov/news-events/news-releases/2018/06/newclinician-screening-tool-available-substance-use.

Drug Brand Names

Amitriptyline • Elavil, Endep
Benztropine • Cogentin
Biperiden • Akineton
Bupropion • Wellbutrin, Zyban
Chlorpromazine • Thorzine
Fluoxetine • Prozac
Haloperidol • Haldol
Olanzapine • Zyprexa
Orphenadrine • Disipal, Norflex
Pregabalin • Lyrica, Lyrica CR
Procyclidine • Kemadrin
Quetiapine • Seroquel
Tianeptine • Coaxil, Stablon
Tranylcypromine • Parnate
Trifluoperazine • Stelazine
Trihexyphenidyl • Artane, Tremin
Venlafaxine • Effexor

The term “dual diagnosis” describes the clinically challenging comorbidity of a substance use disorder (SUD) along with another major mental illness. Based on data from the Epidemiologic Catchment Area study, the lifetime prevalence of SUDs among patients with mental illness is approximately 30%, and is higher among patients with certain mental disorders, such as schizophrenia (47%), bipolar disorder (61%), and antisocial personality disorder (84%).¹ These statistics highlight that addiction is often the rule rather than the exception among those with severe mental illness.¹ Not surprisingly, the combined effects of having an SUD along with another mental illness are uniformly negative (Table 1^2-4).

Based on outcomes research, the core tenets of evidence-based dual-diagnosis treatment include the importance of integrated (rather than parallel) and simultaneous (rather than sequential) care, which means an ideal treatment program includes a unified, multidisciplinary team whose coordinated efforts focus on treating both disorders concurrently.² Evidence-based psychotherapies for addiction, including motivational interviewing, cognitive-behavioral therapy, relapse prevention, contingency management, skills training, and/or case management, are a necessity,^3,5 and must be balanced with rational and appropriate pharmacotherapy targeting both the SUD as well as the other disorder (Table 2^2,3,5-9).

3 ‘Real-world’ clinical challenges

Ideal vs real-world treatment

Treating patients with co-occurring disorders (CODs) within integrated dual-disorder treatment (IDDT) programs sounds straightforward. However, implementing evidence-based “best practice” treatment is a significant challenge in the real world for several reasons. First, individuals with CODs often struggle with poor insight, low motivation to change, and lack of access to health care. According to the Substance Abuse and Mental Health Services Administration (SAMHSA), 52% of individuals with CODs in the U.S. received no treatment at all in 2016.¹⁰ For patients with dual disorders who do seek care, most are not given access to specialty SUD treatment¹⁰ and may instead find themselves treated by psychiatrists with limited SUD training who fail to provide evidence-based psychotherapies and underutilize pharmacotherapies for SUDs.¹¹ In the setting of CODs, the “harm reduction model” can be conflated with therapeutic nihilism, resulting in the neglect of SUD issues, with clinicians expecting patients to seek SUD treatment on their own, through self-help groups such as Alcoholics Anonymous or in other community treatment programs staffed by nonprofessionals that often are not tailored to the unique needs of patients with dual disorders. Psychiatrists working with other mental health professionals who provide psychotherapy for SUDs often do so in parallel rather than in an evidence-based, integrated fashion.

IDDT programs are not widely available. One study found that fewer than 20% of addiction treatment programs and fewer than 10% of mental health programs in the U.S. met criteria for dual diagnosis–capable services.¹² Getting treatment programs to become dual diagnosis–capable is possible, but it is a time-consuming and costly endeavor that, once achieved, requires continuous staff training and programmatic adaptations to interruptions in funding.^13-16 With myriad barriers to the establishment and maintenance of IDDTs, many patients with dual disorders are left without access to the most effective and comprehensive care; as few as 4% of individuals with CODs are treated within integrated programs.¹⁷

Diagnostic dilemmas

Establishing whether or not a patient with an active SUD has another serious mental illness (SMI) is a crucial first step for optimizing treatment, but diagnostic reliability can prove challenging and requires careful clinical assessment (Table 3). As always in psychiatry, accurate diagnosis is limited to careful clinical assessment¹⁸ and, in the case of possible dual disorders, is complicated by the fact that both SUDs as well as non-SUDs can result in the same psychiatric symptoms (eg, insomnia, anxiety, depression, manic behaviors, and psychosis). Clinicians must therefore distinguish between:

• Symptoms of substance intoxication or withdrawal vs independent symptoms of an underlying psychiatric disorder (that persist beyond a month after cessation of intoxication or withdrawal)
• Subclinical symptoms vs threshold mental illness, keeping in mind that some mood and anxiety states can be normal given social situations and stressors (eg, turmoil in relationships, employment difficulties, homelessness, etc.)
• Any mental illness (AMI) vs SMI. The latter is defined by SAMHSA as AMI that substantially interferes with or limits ≥1 major life activities.¹⁰

With these distinctions in mind, data from the 2016 National Survey on Drug Use and Health indicate that dual-diagnosis comorbidity was higher when the thres­hold for mental illness was lower—among the 19 million adults in the U.S. with SUDs, the past-year prevalence was 43% for AMI and 14% for SMI.¹⁰ Looking at substance-induced disorders vs “independent” disorders, the 2001-2002 National Epidemiologic Survey on Alcohol and Related Conditions found that for individuals with SUDs, the past-year prevalence of an independent mood or anxiety disorder was 35% and 26%, respectively.¹⁹ Taken together, these findings illustrate the substantial rate of dual-diagnosis comorbidity, the diagnostic heterogeneity and range of severity of CODs,²⁰ and the potential for both false negatives (eg, diagnosing a substance-induced syndrome when in fact a patient has an underlying disorder) and false positives (diagnosing a full-blown mental illness when symptoms are subclinical or substance-induced) when performing diagnostic assessments in the setting of known SUDs.

Continue to: False positives are more likely...

False positives are more likely when patients seeking treatment for non-SUDs don’t disclose active drug use, even when asked. Both patients and their treating clinicians may also be prone to underestimating the significant potential for morbidity associated with SUDs, such that substance-induced symptoms may be misattributed to a dual disorder. Diagnostic questioning and thorough chart review that includes careful assessment of whether psychiatric symptoms preceded the onset of substance use, and whether they persisted in the setting of extended sobriety, is therefore paramount for minimizing false positives when assessing for dual diagnoses.^18,21 Likewise, random urine toxicology testing can be invaluable in verifying claims regarding sobriety.

Another factor that can complicate diagnosis is that there are often considerable secondary gains (eg, disability income, hospitalization, housing, access to prescription medications, and mitigation of the blame and stigma associated with addiction) associated with having a dual disorder as opposed to having “just” a SUD. As a result, for some patients, obtaining a non-SUD diagnosis can be highly incentivized.^22,23 Clinicians must therefore be savvy about the high potential for malingering, embellishment, and mislabeling of symptoms when conducting diagnostic interviews. For example, in assessing for psychosis, the frequent endorsement of “hearing voices” in patients with SUDs often results in a diagnosis of schizophrenia or unspecified psychotic disorder,²² despite the fact that this symptom can occur during substance intoxication and withdrawal, is well documented among people without mental illness as well as those with non-psychotic disorders,²⁴ and can resolve without medications or with non-antipsychotic pharmacotherapy.²⁵

When assessing for dual disorders, diagnostic false positives and false negatives can both contribute to inappropriate treatment and unrealistic expectations for recovery, and therefore underscore the importance of careful diagnostic assessment. Even with diligent assessment, however, diagnostic clarity can prove elusive due to inadequate sobriety, inconsistent reporting, and poor memory.²⁶ Therefore, for patients with known SUDs but diagnostic uncertainty about a dual disorder, the work-up should include a trial of prospective observation, with completion of appropriate detoxification, throughout a 1-month period of sobriety and in the absence of psychiatric medications, to determine if there are persistent symptoms that would justify a dual diagnosis. In research settings, such observations have revealed that most of depressive symptoms among alcoholics who present for substance abuse treatment resolve after a month of abstinence.²⁷ A similar time course for resolution has been noted for anxiety, distress, fatigue, and depressive symptoms among individuals with cocaine dependence.²⁸ These findings support the guideline established in DSM-IV that symptoms persisting beyond a month of sobriety “should be considered to be manifestations of an independent, non-substance-induced mental disorder,”²⁹ while symptoms occurring within that month may well be substance-induced. Unfortunately, in real-world clinical practice, and particularly in outpatient settings, it can be quite difficult to achieve the requisite period of sobriety for reliable diagnosis, and patients are often prematurely prescribed medications (eg, an antidepressant, antipsychotic, or mood stabilizer) that can confound the cause of symptomatic resolution. Such prescriptions are driven by compelling pressures from patients to relieve their acute suffering, as well as the predilection of some clinicians to give patients “the benefit of doubt” in assessing for dual diagnoses. However, whether an inappropriate diagnosis or a prescription for an unnecessary medication represents a benefit is debatable at best.

Pharmacotherapy

A third real-world challenge in managing patients with dual disorders involves optimizing pharmacotherapy. Unfortunately, because patients with SUDs often are excluded from clinical trials, evidence-based guidance for patients with dual disorders is lacking. In addition, medications for both CODs often remain inaccessible to patients with dual disorders for 3 reasons:

• SUDs negatively impact medication adherence among patients with dual disorders, who sometimes point out that “it says right here on the bottle not to take this medication with drugs or alcohol!”
• Some self-help groups still espouse blanket opposition of any “psychotropic” medications, even when clearly indicated for patients with COD. Groups that recognize the importance of pharmacotherapy, such as Dual Diagnosis Anonymous (DDA), have emerged, but are not yet widely available.³⁰
• Although there are increasing options for FDA-approved medications for SUDs, they are limited to the treatment of alcohol, opioid, and nicotine use disorders³¹; are often restricted due to hospital and health insurance formularies³²; and remain underprescribed for patients with dual disorders.¹¹

Continue to: Although underutilization of pharmacotherapy is...

Although underutilization of pharmacotherapy is a pitfall to be avoided in the treatment of patients with dual disorders, medication overutilization can be just as problematic. Patients with dual disorders are sometimes singularly focused on resolving acute anxiety, depression, or psychosis at the expense of working towards sobriety.³³ Although the “self-medication hypothesis” is frequently invoked by patients and clinicians alike to suggest that substance use occurs in the service of “treating” underlying disorders,³⁴ this theory has not been well supported in studies.^35-37 Some patients may pledge dedication to abstinence, but still pressure physicians for a pharmacologic solution to their suffering. With expanding legalization of cannabis for both recreational and medical purposes, patients are increasingly seeking doctors’ recommendations for “medical marijuana” for a wide range of complaints, despite the fact that data supporting a therapeutic role for cannabis in the treatment of mental illness is sparse,³⁸ whereas the potential harm in terms of either causing or worsening psychosis is well established.^39,40 Clinicians must be knowledgeable about the abuse potential of prescribed medications, ranging from sleep aids, analgesics, and muscle relaxants to antidepressants and antipsychotics, while also being mindful of the psychological meaningfulness of seeking, prescribing, and not prescribing medications.⁴¹

Although the simultaneous treatment of patients with dual disorders that includes pharmacotherapy for both SUDs and CODs is vital for optimizing clinical outcomes, clinicians should strive for diagnostic accuracy and use medications judiciously. In addition, although pharmacotherapy often is necessary to deliver evidence-based treatment for patients with dual disorders, it is inadequate as standalone treatment and should be administered along with psychosocial interventions within an integrated, multidisciplinary treatment setting.

The keys to optimal outcomes

The treatment of patients with dual disorders can be challenging, to say the least. Ideal, evidence-based therapy in the form of an IDDT program can be difficult for clinicians to implement and for patients to access. Best efforts to perform meticulous clinical assessment to clarify diagnoses, use pharmacotherapy judiciously, work collaboratively in a multidisciplinary setting, and optimize treatment given available resources are keys to clinical success.

Bottom Line

Ideal treatment of patients with dual disorders consists of simultaneous, integrated interventions delivered by a multidisciplinary team. However, in the real world, limited resources, diagnostic challenges, and both over- and underutilization of pharmacotherapy often hamper optimal treatment.

Related Resources

• Substance Abuse and Mental Health Services Administration. Co-occurring disorders. https://www.samhsa.gov/disorders/co-occurring.
• National Alliance on Mental Illness. Dual diagnosis. https://www.nami.org/Learn-More/Mental-Health-Conditions/Related-Conditions/Dual-Diagnosis.
• Foundations Recovery Network. Dual diagnosis support groups. http://www.dualdiagnosis.org/resource/ddrn/self-helpsupport-groups/support-groups.
• Substance Abuse and Mental Health Services Administration. Integrated treatment for co-occurring disorders evidence-based practices (EBP) KIT. https://store.samhsa.gov/product/Integrated-Treatment-for-Co-Occurring-Disorders-Evidence-Based-Practices-EBP-KIT/SMA08-4367.
• Substance Abuse and Mental Health Services Administration. Substance abuse treatment for persons with co-occurring disorders. https://store.samhsa.gov/shin/content/SMA13-3992/SMA13-3992.pdf.

The term “dual diagnosis” describes the clinically challenging comorbidity of a substance use disorder (SUD) along with another major mental illness. Based on data from the Epidemiologic Catchment Area study, the lifetime prevalence of SUDs among patients with mental illness is approximately 30%, and is higher among patients with certain mental disorders, such as schizophrenia (47%), bipolar disorder (61%), and antisocial personality disorder (84%).¹ These statistics highlight that addiction is often the rule rather than the exception among those with severe mental illness.¹ Not surprisingly, the combined effects of having an SUD along with another mental illness are uniformly negative (Table 1^2-4).

Based on outcomes research, the core tenets of evidence-based dual-diagnosis treatment include the importance of integrated (rather than parallel) and simultaneous (rather than sequential) care, which means an ideal treatment program includes a unified, multidisciplinary team whose coordinated efforts focus on treating both disorders concurrently.² Evidence-based psychotherapies for addiction, including motivational interviewing, cognitive-behavioral therapy, relapse prevention, contingency management, skills training, and/or case management, are a necessity,^3,5 and must be balanced with rational and appropriate pharmacotherapy targeting both the SUD as well as the other disorder (Table 2^2,3,5-9).

3 ‘Real-world’ clinical challenges

Ideal vs real-world treatment

Treating patients with co-occurring disorders (CODs) within integrated dual-disorder treatment (IDDT) programs sounds straightforward. However, implementing evidence-based “best practice” treatment is a significant challenge in the real world for several reasons. First, individuals with CODs often struggle with poor insight, low motivation to change, and lack of access to health care. According to the Substance Abuse and Mental Health Services Administration (SAMHSA), 52% of individuals with CODs in the U.S. received no treatment at all in 2016.¹⁰ For patients with dual disorders who do seek care, most are not given access to specialty SUD treatment¹⁰ and may instead find themselves treated by psychiatrists with limited SUD training who fail to provide evidence-based psychotherapies and underutilize pharmacotherapies for SUDs.¹¹ In the setting of CODs, the “harm reduction model” can be conflated with therapeutic nihilism, resulting in the neglect of SUD issues, with clinicians expecting patients to seek SUD treatment on their own, through self-help groups such as Alcoholics Anonymous or in other community treatment programs staffed by nonprofessionals that often are not tailored to the unique needs of patients with dual disorders. Psychiatrists working with other mental health professionals who provide psychotherapy for SUDs often do so in parallel rather than in an evidence-based, integrated fashion.

IDDT programs are not widely available. One study found that fewer than 20% of addiction treatment programs and fewer than 10% of mental health programs in the U.S. met criteria for dual diagnosis–capable services.¹² Getting treatment programs to become dual diagnosis–capable is possible, but it is a time-consuming and costly endeavor that, once achieved, requires continuous staff training and programmatic adaptations to interruptions in funding.^13-16 With myriad barriers to the establishment and maintenance of IDDTs, many patients with dual disorders are left without access to the most effective and comprehensive care; as few as 4% of individuals with CODs are treated within integrated programs.¹⁷

Diagnostic dilemmas

Establishing whether or not a patient with an active SUD has another serious mental illness (SMI) is a crucial first step for optimizing treatment, but diagnostic reliability can prove challenging and requires careful clinical assessment (Table 3). As always in psychiatry, accurate diagnosis is limited to careful clinical assessment¹⁸ and, in the case of possible dual disorders, is complicated by the fact that both SUDs as well as non-SUDs can result in the same psychiatric symptoms (eg, insomnia, anxiety, depression, manic behaviors, and psychosis). Clinicians must therefore distinguish between:

• Symptoms of substance intoxication or withdrawal vs independent symptoms of an underlying psychiatric disorder (that persist beyond a month after cessation of intoxication or withdrawal)
• Subclinical symptoms vs threshold mental illness, keeping in mind that some mood and anxiety states can be normal given social situations and stressors (eg, turmoil in relationships, employment difficulties, homelessness, etc.)
• Any mental illness (AMI) vs SMI. The latter is defined by SAMHSA as AMI that substantially interferes with or limits ≥1 major life activities.¹⁰

With these distinctions in mind, data from the 2016 National Survey on Drug Use and Health indicate that dual-diagnosis comorbidity was higher when the thres­hold for mental illness was lower—among the 19 million adults in the U.S. with SUDs, the past-year prevalence was 43% for AMI and 14% for SMI.¹⁰ Looking at substance-induced disorders vs “independent” disorders, the 2001-2002 National Epidemiologic Survey on Alcohol and Related Conditions found that for individuals with SUDs, the past-year prevalence of an independent mood or anxiety disorder was 35% and 26%, respectively.¹⁹ Taken together, these findings illustrate the substantial rate of dual-diagnosis comorbidity, the diagnostic heterogeneity and range of severity of CODs,²⁰ and the potential for both false negatives (eg, diagnosing a substance-induced syndrome when in fact a patient has an underlying disorder) and false positives (diagnosing a full-blown mental illness when symptoms are subclinical or substance-induced) when performing diagnostic assessments in the setting of known SUDs.

Continue to: False positives are more likely...

False positives are more likely when patients seeking treatment for non-SUDs don’t disclose active drug use, even when asked. Both patients and their treating clinicians may also be prone to underestimating the significant potential for morbidity associated with SUDs, such that substance-induced symptoms may be misattributed to a dual disorder. Diagnostic questioning and thorough chart review that includes careful assessment of whether psychiatric symptoms preceded the onset of substance use, and whether they persisted in the setting of extended sobriety, is therefore paramount for minimizing false positives when assessing for dual diagnoses.^18,21 Likewise, random urine toxicology testing can be invaluable in verifying claims regarding sobriety.

Another factor that can complicate diagnosis is that there are often considerable secondary gains (eg, disability income, hospitalization, housing, access to prescription medications, and mitigation of the blame and stigma associated with addiction) associated with having a dual disorder as opposed to having “just” a SUD. As a result, for some patients, obtaining a non-SUD diagnosis can be highly incentivized.^22,23 Clinicians must therefore be savvy about the high potential for malingering, embellishment, and mislabeling of symptoms when conducting diagnostic interviews. For example, in assessing for psychosis, the frequent endorsement of “hearing voices” in patients with SUDs often results in a diagnosis of schizophrenia or unspecified psychotic disorder,²² despite the fact that this symptom can occur during substance intoxication and withdrawal, is well documented among people without mental illness as well as those with non-psychotic disorders,²⁴ and can resolve without medications or with non-antipsychotic pharmacotherapy.²⁵

When assessing for dual disorders, diagnostic false positives and false negatives can both contribute to inappropriate treatment and unrealistic expectations for recovery, and therefore underscore the importance of careful diagnostic assessment. Even with diligent assessment, however, diagnostic clarity can prove elusive due to inadequate sobriety, inconsistent reporting, and poor memory.²⁶ Therefore, for patients with known SUDs but diagnostic uncertainty about a dual disorder, the work-up should include a trial of prospective observation, with completion of appropriate detoxification, throughout a 1-month period of sobriety and in the absence of psychiatric medications, to determine if there are persistent symptoms that would justify a dual diagnosis. In research settings, such observations have revealed that most of depressive symptoms among alcoholics who present for substance abuse treatment resolve after a month of abstinence.²⁷ A similar time course for resolution has been noted for anxiety, distress, fatigue, and depressive symptoms among individuals with cocaine dependence.²⁸ These findings support the guideline established in DSM-IV that symptoms persisting beyond a month of sobriety “should be considered to be manifestations of an independent, non-substance-induced mental disorder,”²⁹ while symptoms occurring within that month may well be substance-induced. Unfortunately, in real-world clinical practice, and particularly in outpatient settings, it can be quite difficult to achieve the requisite period of sobriety for reliable diagnosis, and patients are often prematurely prescribed medications (eg, an antidepressant, antipsychotic, or mood stabilizer) that can confound the cause of symptomatic resolution. Such prescriptions are driven by compelling pressures from patients to relieve their acute suffering, as well as the predilection of some clinicians to give patients “the benefit of doubt” in assessing for dual diagnoses. However, whether an inappropriate diagnosis or a prescription for an unnecessary medication represents a benefit is debatable at best.

Pharmacotherapy

A third real-world challenge in managing patients with dual disorders involves optimizing pharmacotherapy. Unfortunately, because patients with SUDs often are excluded from clinical trials, evidence-based guidance for patients with dual disorders is lacking. In addition, medications for both CODs often remain inaccessible to patients with dual disorders for 3 reasons:

• SUDs negatively impact medication adherence among patients with dual disorders, who sometimes point out that “it says right here on the bottle not to take this medication with drugs or alcohol!”
• Some self-help groups still espouse blanket opposition of any “psychotropic” medications, even when clearly indicated for patients with COD. Groups that recognize the importance of pharmacotherapy, such as Dual Diagnosis Anonymous (DDA), have emerged, but are not yet widely available.³⁰
• Although there are increasing options for FDA-approved medications for SUDs, they are limited to the treatment of alcohol, opioid, and nicotine use disorders³¹; are often restricted due to hospital and health insurance formularies³²; and remain underprescribed for patients with dual disorders.¹¹

Continue to: Although underutilization of pharmacotherapy is...

Although underutilization of pharmacotherapy is a pitfall to be avoided in the treatment of patients with dual disorders, medication overutilization can be just as problematic. Patients with dual disorders are sometimes singularly focused on resolving acute anxiety, depression, or psychosis at the expense of working towards sobriety.³³ Although the “self-medication hypothesis” is frequently invoked by patients and clinicians alike to suggest that substance use occurs in the service of “treating” underlying disorders,³⁴ this theory has not been well supported in studies.^35-37 Some patients may pledge dedication to abstinence, but still pressure physicians for a pharmacologic solution to their suffering. With expanding legalization of cannabis for both recreational and medical purposes, patients are increasingly seeking doctors’ recommendations for “medical marijuana” for a wide range of complaints, despite the fact that data supporting a therapeutic role for cannabis in the treatment of mental illness is sparse,³⁸ whereas the potential harm in terms of either causing or worsening psychosis is well established.^39,40 Clinicians must be knowledgeable about the abuse potential of prescribed medications, ranging from sleep aids, analgesics, and muscle relaxants to antidepressants and antipsychotics, while also being mindful of the psychological meaningfulness of seeking, prescribing, and not prescribing medications.⁴¹

Although the simultaneous treatment of patients with dual disorders that includes pharmacotherapy for both SUDs and CODs is vital for optimizing clinical outcomes, clinicians should strive for diagnostic accuracy and use medications judiciously. In addition, although pharmacotherapy often is necessary to deliver evidence-based treatment for patients with dual disorders, it is inadequate as standalone treatment and should be administered along with psychosocial interventions within an integrated, multidisciplinary treatment setting.

The keys to optimal outcomes

The treatment of patients with dual disorders can be challenging, to say the least. Ideal, evidence-based therapy in the form of an IDDT program can be difficult for clinicians to implement and for patients to access. Best efforts to perform meticulous clinical assessment to clarify diagnoses, use pharmacotherapy judiciously, work collaboratively in a multidisciplinary setting, and optimize treatment given available resources are keys to clinical success.

Bottom Line

Ideal treatment of patients with dual disorders consists of simultaneous, integrated interventions delivered by a multidisciplinary team. However, in the real world, limited resources, diagnostic challenges, and both over- and underutilization of pharmacotherapy often hamper optimal treatment.

Related Resources

• Substance Abuse and Mental Health Services Administration. Co-occurring disorders. https://www.samhsa.gov/disorders/co-occurring.
• National Alliance on Mental Illness. Dual diagnosis. https://www.nami.org/Learn-More/Mental-Health-Conditions/Related-Conditions/Dual-Diagnosis.
• Foundations Recovery Network. Dual diagnosis support groups. http://www.dualdiagnosis.org/resource/ddrn/self-helpsupport-groups/support-groups.
• Substance Abuse and Mental Health Services Administration. Integrated treatment for co-occurring disorders evidence-based practices (EBP) KIT. https://store.samhsa.gov/product/Integrated-Treatment-for-Co-Occurring-Disorders-Evidence-Based-Practices-EBP-KIT/SMA08-4367.
• Substance Abuse and Mental Health Services Administration. Substance abuse treatment for persons with co-occurring disorders. https://store.samhsa.gov/shin/content/SMA13-3992/SMA13-3992.pdf.

The term “dual diagnosis” describes the clinically challenging comorbidity of a substance use disorder (SUD) along with another major mental illness. Based on data from the Epidemiologic Catchment Area study, the lifetime prevalence of SUDs among patients with mental illness is approximately 30%, and is higher among patients with certain mental disorders, such as schizophrenia (47%), bipolar disorder (61%), and antisocial personality disorder (84%).¹ These statistics highlight that addiction is often the rule rather than the exception among those with severe mental illness.¹ Not surprisingly, the combined effects of having an SUD along with another mental illness are uniformly negative (Table 1^2-4).

Based on outcomes research, the core tenets of evidence-based dual-diagnosis treatment include the importance of integrated (rather than parallel) and simultaneous (rather than sequential) care, which means an ideal treatment program includes a unified, multidisciplinary team whose coordinated efforts focus on treating both disorders concurrently.² Evidence-based psychotherapies for addiction, including motivational interviewing, cognitive-behavioral therapy, relapse prevention, contingency management, skills training, and/or case management, are a necessity,^3,5 and must be balanced with rational and appropriate pharmacotherapy targeting both the SUD as well as the other disorder (Table 2^2,3,5-9).

3 ‘Real-world’ clinical challenges

Ideal vs real-world treatment

Treating patients with co-occurring disorders (CODs) within integrated dual-disorder treatment (IDDT) programs sounds straightforward. However, implementing evidence-based “best practice” treatment is a significant challenge in the real world for several reasons. First, individuals with CODs often struggle with poor insight, low motivation to change, and lack of access to health care. According to the Substance Abuse and Mental Health Services Administration (SAMHSA), 52% of individuals with CODs in the U.S. received no treatment at all in 2016.¹⁰ For patients with dual disorders who do seek care, most are not given access to specialty SUD treatment¹⁰ and may instead find themselves treated by psychiatrists with limited SUD training who fail to provide evidence-based psychotherapies and underutilize pharmacotherapies for SUDs.¹¹ In the setting of CODs, the “harm reduction model” can be conflated with therapeutic nihilism, resulting in the neglect of SUD issues, with clinicians expecting patients to seek SUD treatment on their own, through self-help groups such as Alcoholics Anonymous or in other community treatment programs staffed by nonprofessionals that often are not tailored to the unique needs of patients with dual disorders. Psychiatrists working with other mental health professionals who provide psychotherapy for SUDs often do so in parallel rather than in an evidence-based, integrated fashion.

IDDT programs are not widely available. One study found that fewer than 20% of addiction treatment programs and fewer than 10% of mental health programs in the U.S. met criteria for dual diagnosis–capable services.¹² Getting treatment programs to become dual diagnosis–capable is possible, but it is a time-consuming and costly endeavor that, once achieved, requires continuous staff training and programmatic adaptations to interruptions in funding.^13-16 With myriad barriers to the establishment and maintenance of IDDTs, many patients with dual disorders are left without access to the most effective and comprehensive care; as few as 4% of individuals with CODs are treated within integrated programs.¹⁷

Diagnostic dilemmas

Establishing whether or not a patient with an active SUD has another serious mental illness (SMI) is a crucial first step for optimizing treatment, but diagnostic reliability can prove challenging and requires careful clinical assessment (Table 3). As always in psychiatry, accurate diagnosis is limited to careful clinical assessment¹⁸ and, in the case of possible dual disorders, is complicated by the fact that both SUDs as well as non-SUDs can result in the same psychiatric symptoms (eg, insomnia, anxiety, depression, manic behaviors, and psychosis). Clinicians must therefore distinguish between:

• Symptoms of substance intoxication or withdrawal vs independent symptoms of an underlying psychiatric disorder (that persist beyond a month after cessation of intoxication or withdrawal)
• Subclinical symptoms vs threshold mental illness, keeping in mind that some mood and anxiety states can be normal given social situations and stressors (eg, turmoil in relationships, employment difficulties, homelessness, etc.)
• Any mental illness (AMI) vs SMI. The latter is defined by SAMHSA as AMI that substantially interferes with or limits ≥1 major life activities.¹⁰

With these distinctions in mind, data from the 2016 National Survey on Drug Use and Health indicate that dual-diagnosis comorbidity was higher when the thres­hold for mental illness was lower—among the 19 million adults in the U.S. with SUDs, the past-year prevalence was 43% for AMI and 14% for SMI.¹⁰ Looking at substance-induced disorders vs “independent” disorders, the 2001-2002 National Epidemiologic Survey on Alcohol and Related Conditions found that for individuals with SUDs, the past-year prevalence of an independent mood or anxiety disorder was 35% and 26%, respectively.¹⁹ Taken together, these findings illustrate the substantial rate of dual-diagnosis comorbidity, the diagnostic heterogeneity and range of severity of CODs,²⁰ and the potential for both false negatives (eg, diagnosing a substance-induced syndrome when in fact a patient has an underlying disorder) and false positives (diagnosing a full-blown mental illness when symptoms are subclinical or substance-induced) when performing diagnostic assessments in the setting of known SUDs.

Continue to: False positives are more likely...

False positives are more likely when patients seeking treatment for non-SUDs don’t disclose active drug use, even when asked. Both patients and their treating clinicians may also be prone to underestimating the significant potential for morbidity associated with SUDs, such that substance-induced symptoms may be misattributed to a dual disorder. Diagnostic questioning and thorough chart review that includes careful assessment of whether psychiatric symptoms preceded the onset of substance use, and whether they persisted in the setting of extended sobriety, is therefore paramount for minimizing false positives when assessing for dual diagnoses.^18,21 Likewise, random urine toxicology testing can be invaluable in verifying claims regarding sobriety.

Another factor that can complicate diagnosis is that there are often considerable secondary gains (eg, disability income, hospitalization, housing, access to prescription medications, and mitigation of the blame and stigma associated with addiction) associated with having a dual disorder as opposed to having “just” a SUD. As a result, for some patients, obtaining a non-SUD diagnosis can be highly incentivized.^22,23 Clinicians must therefore be savvy about the high potential for malingering, embellishment, and mislabeling of symptoms when conducting diagnostic interviews. For example, in assessing for psychosis, the frequent endorsement of “hearing voices” in patients with SUDs often results in a diagnosis of schizophrenia or unspecified psychotic disorder,²² despite the fact that this symptom can occur during substance intoxication and withdrawal, is well documented among people without mental illness as well as those with non-psychotic disorders,²⁴ and can resolve without medications or with non-antipsychotic pharmacotherapy.²⁵

When assessing for dual disorders, diagnostic false positives and false negatives can both contribute to inappropriate treatment and unrealistic expectations for recovery, and therefore underscore the importance of careful diagnostic assessment. Even with diligent assessment, however, diagnostic clarity can prove elusive due to inadequate sobriety, inconsistent reporting, and poor memory.²⁶ Therefore, for patients with known SUDs but diagnostic uncertainty about a dual disorder, the work-up should include a trial of prospective observation, with completion of appropriate detoxification, throughout a 1-month period of sobriety and in the absence of psychiatric medications, to determine if there are persistent symptoms that would justify a dual diagnosis. In research settings, such observations have revealed that most of depressive symptoms among alcoholics who present for substance abuse treatment resolve after a month of abstinence.²⁷ A similar time course for resolution has been noted for anxiety, distress, fatigue, and depressive symptoms among individuals with cocaine dependence.²⁸ These findings support the guideline established in DSM-IV that symptoms persisting beyond a month of sobriety “should be considered to be manifestations of an independent, non-substance-induced mental disorder,”²⁹ while symptoms occurring within that month may well be substance-induced. Unfortunately, in real-world clinical practice, and particularly in outpatient settings, it can be quite difficult to achieve the requisite period of sobriety for reliable diagnosis, and patients are often prematurely prescribed medications (eg, an antidepressant, antipsychotic, or mood stabilizer) that can confound the cause of symptomatic resolution. Such prescriptions are driven by compelling pressures from patients to relieve their acute suffering, as well as the predilection of some clinicians to give patients “the benefit of doubt” in assessing for dual diagnoses. However, whether an inappropriate diagnosis or a prescription for an unnecessary medication represents a benefit is debatable at best.

Pharmacotherapy

A third real-world challenge in managing patients with dual disorders involves optimizing pharmacotherapy. Unfortunately, because patients with SUDs often are excluded from clinical trials, evidence-based guidance for patients with dual disorders is lacking. In addition, medications for both CODs often remain inaccessible to patients with dual disorders for 3 reasons:

• SUDs negatively impact medication adherence among patients with dual disorders, who sometimes point out that “it says right here on the bottle not to take this medication with drugs or alcohol!”
• Some self-help groups still espouse blanket opposition of any “psychotropic” medications, even when clearly indicated for patients with COD. Groups that recognize the importance of pharmacotherapy, such as Dual Diagnosis Anonymous (DDA), have emerged, but are not yet widely available.³⁰
• Although there are increasing options for FDA-approved medications for SUDs, they are limited to the treatment of alcohol, opioid, and nicotine use disorders³¹; are often restricted due to hospital and health insurance formularies³²; and remain underprescribed for patients with dual disorders.¹¹

Continue to: Although underutilization of pharmacotherapy is...

Although underutilization of pharmacotherapy is a pitfall to be avoided in the treatment of patients with dual disorders, medication overutilization can be just as problematic. Patients with dual disorders are sometimes singularly focused on resolving acute anxiety, depression, or psychosis at the expense of working towards sobriety.³³ Although the “self-medication hypothesis” is frequently invoked by patients and clinicians alike to suggest that substance use occurs in the service of “treating” underlying disorders,³⁴ this theory has not been well supported in studies.^35-37 Some patients may pledge dedication to abstinence, but still pressure physicians for a pharmacologic solution to their suffering. With expanding legalization of cannabis for both recreational and medical purposes, patients are increasingly seeking doctors’ recommendations for “medical marijuana” for a wide range of complaints, despite the fact that data supporting a therapeutic role for cannabis in the treatment of mental illness is sparse,³⁸ whereas the potential harm in terms of either causing or worsening psychosis is well established.^39,40 Clinicians must be knowledgeable about the abuse potential of prescribed medications, ranging from sleep aids, analgesics, and muscle relaxants to antidepressants and antipsychotics, while also being mindful of the psychological meaningfulness of seeking, prescribing, and not prescribing medications.⁴¹

Although the simultaneous treatment of patients with dual disorders that includes pharmacotherapy for both SUDs and CODs is vital for optimizing clinical outcomes, clinicians should strive for diagnostic accuracy and use medications judiciously. In addition, although pharmacotherapy often is necessary to deliver evidence-based treatment for patients with dual disorders, it is inadequate as standalone treatment and should be administered along with psychosocial interventions within an integrated, multidisciplinary treatment setting.

The keys to optimal outcomes

The treatment of patients with dual disorders can be challenging, to say the least. Ideal, evidence-based therapy in the form of an IDDT program can be difficult for clinicians to implement and for patients to access. Best efforts to perform meticulous clinical assessment to clarify diagnoses, use pharmacotherapy judiciously, work collaboratively in a multidisciplinary setting, and optimize treatment given available resources are keys to clinical success.

Bottom Line

Ideal treatment of patients with dual disorders consists of simultaneous, integrated interventions delivered by a multidisciplinary team. However, in the real world, limited resources, diagnostic challenges, and both over- and underutilization of pharmacotherapy often hamper optimal treatment.

Related Resources

• Substance Abuse and Mental Health Services Administration. Co-occurring disorders. https://www.samhsa.gov/disorders/co-occurring.
• National Alliance on Mental Illness. Dual diagnosis. https://www.nami.org/Learn-More/Mental-Health-Conditions/Related-Conditions/Dual-Diagnosis.
• Foundations Recovery Network. Dual diagnosis support groups. http://www.dualdiagnosis.org/resource/ddrn/self-helpsupport-groups/support-groups.
• Substance Abuse and Mental Health Services Administration. Integrated treatment for co-occurring disorders evidence-based practices (EBP) KIT. https://store.samhsa.gov/product/Integrated-Treatment-for-Co-Occurring-Disorders-Evidence-Based-Practices-EBP-KIT/SMA08-4367.
• Substance Abuse and Mental Health Services Administration. Substance abuse treatment for persons with co-occurring disorders. https://store.samhsa.gov/shin/content/SMA13-3992/SMA13-3992.pdf.

In the United States, the average potency of Cannabis has increased significantly over the past few decades in response to consumer demand and policies in some states that have legalized marijuana for medicinal and recreational purposes.¹ Whereas the delta-9-tetrahydrocannabinol (THC) content of “street” marijuana was <1% in the 1970s and 4% in the 1990s, by 2012, analyses of Cannabis samples seized by law enforcement agencies documented a rise in average THC potency to >12%.^1-3

Although this increase in potency has been overstated in the media because studies did not control for the effects of changes in sampling methods on freshness, it is estimated that Cannabis potency increased 7-fold from 1970 to 2010.³ Also, Cannabis preparations such as hashish and hash oil extracts containing THC well above average—from 35% to 90% THC—are now more widely available. In states where marijuana has been legalized, high-potency Cannabis (HPC) in the form of “edibles” (eg, marijuana added to baked goods, candy, or drinks) and hash oil extracts (Table 1)^4-13 can be readily obtained from dispensaries or even at local farmers’ markets.

The potency of Cannabis, typically defined as the percentage of THC, its chief psychoactive component, varies depending on the genetic strain of the plant, cultivation techniques, and methods of processing and storage. For example, relative to “average marijuana,” hemp (Cannabis bred for industrial purposes) has very little THC, while sinsemilla (flowering buds from unpollinated female plants), hashish (Cannabis resin), and extracted hash oil contain increasing amounts of THC (Table 2).^1,2

Cannabis use disorder

Recent data indicate that the prevalence of Cannabis use disorders (eg, abuse and dependence) in the United States is approximately 3% among the general population and >30% among Cannabis users.¹⁶ The availability of increasingly potent forms of Cannabis has been cited as a possible explanation for this rise, despite no change in the prevalence of overall marijuana use between 1991 to 1992 and 2001 to 2002.¹⁷ However, while the prevalence of marijuana use disorders has continued to rise—nearly doubling from 2001 to 2002 to 2012 to 2013—this latest increase occurred with a significant increase in overall marijuana use, such that the actual rate of Cannabis use disorders among users seems to have plateaued, despite the continued rise in marijuana potency.¹⁶ This discrepancy could be explained if Cannabis users cut back past a specific threshold of increasing potency. However, 2 studies have called into question how effective such titration efforts might be in practice. In one study, Cannabis users who preferred more potent Cannabis inhaled lower volumes of smoke, but did not fully compensate for the increased potency, such that use of HPC still resulted in greater THC exposure.¹⁸ Another study found that HPC users rolled less marijuana into their joints but not enough to mitigate the impact of greater potency.¹⁹ Therefore, it appears that HPC users typically expose themselves to greater amounts of THC, which could place them at higher risk of addiction.

Although a causal association between increasing Cannabis potency and the rate of substance use disorders among users remains unclear based on epidemiologic studies from the United States, a recent study from the United Kingdom examined the impact of Cannabis potency on dependence.²⁰ This cross-sectional survey found that, although HPC was preferred by users and was rated as offering the “best high,” its use was associated with increasing severity of dependence, especially among young people. The limited available evidence supports a greater risk of Cannabis use disorders with increasing potency.

Psychosis

Based on longitudinal studies published over the past 30 years, it is clear that using Cannabis at a young age (age <15 to 18) increases the risk of developing a psychotic disorder.²¹ This association appears to be dose-dependent, with studies consistently demonstrating that psychosis risk increases with greater frequency of Cannabis use.²² The accumulated evidence to date is strong enough to view the psychotic potential of Cannabis as a significant public health concern.²¹

If risk of psychosis is proportional to the amount of Cannabis used as measured by frequency, it follows that this risk might be affected similarly by Cannabis potency. In another paper, I discussed the potential for greater risk of psychosis in the context of medical marijuana and synthetic cannabinoids.²³ My colleagues and I also have published case reports describing emerging psychosis among regular Cannabis users after escalating to higher potency medical marijuana²⁴ and a hyperconcentrated form of hash oil known as Cannabis “wax” or “dabs” that contains as much as 90% THC.⁴ Preliminary anecdotal evidence supports the plausibility of HPC being more psychotoxic than less potent forms.

Several studies from a research group in the United Kingdom, where sinsemilla has increasingly dominated the drug market, likewise have reported that the use of HPC is associated with a greater risk of psychosis. The first of these studies, published in 2009, found that adults hospitalized for first-episode psychosis were more likely to have used HPC than healthy controls.²⁵ Among Cannabis users, HPC use was associated with a 7-fold increased risk of psychosis, with daily HPC use associated with a 12-fold increased risk.

Based on a larger dataset, a second study reported that high-potency, but not low-potency, Cannabis increased the risk of first-episode psychosis with increasing frequency of use.²⁶ Daily users of HPC had a 5-fold higher risk of psychosis compared with those that had never used Cannabis. A third study reported that HPC use and daily Cannabis use were independently associated with an earlier onset of first-episode psychosis, with daily HPC users developing first-episode psychosis an average of 6 years earlier than non-Cannabis users.²⁷ Finally, a prospective study following patients with first-episode psychosis over 2 years found that the greatest risk of relapse—defined by hospital admission caused by exacerbation of psychotic symptoms—was found among self-reported daily users of HPC, while the lowest risk was among those who stopped using Cannabis after their initial psychotic episode.²⁸

The findings from these 4 studies suggest that the increased risk of psychosis with Cannabis is proportional to overall exposure, determined by both frequency of use and Cannabis potency.

Cognition

There is little doubt that using Cannabis can impair cognition acutely, “after all, this is the basic reason for its recreational use,” as one author wrote.²⁹ As with psychosis, the available evidence indicates that the degree of cognitive impairment is related to the frequency and duration of Cannabis use as well as age of onset of use.^30,31

Few studies have assessed cognitive functioning in relation to Cannabis potency with most only examining the effects of relatively low-potency Cannabis with inconsistent results. For example, 2 studies compared cognitive performance in individuals smoking Cannabis with 1.8% and 3.9% THC. One study found that using higher potency Cannabis resulted in prolonged time needed to complete certain cognitive tasks,³² whereas the other found greater impairment in performance on a decision-making task at both potencies compared with non-users but no differences between the 2 dosages.³³ Detecting significant differences may be difficult within the narrow range of low Cannabis potency studied where any findings have limited applicability in the context of today’s Cannabis with much higher THC content.

To date, only 1 study has assessed cognition at higher Cannabis potencies, comparing Cannabis with 4% THC to 13% THC.³⁴ Cognitive impairments increased with higher potency, especially in tasks that measured motor control and executive functioning. Therefore it appears that higher potency Cannabis use is associated with greater acute cognitive impairment.

The longer-term effects on cognition are less clear, with conflicting evidence about whether Cannabis use can result in residual cognitive impairment despite abstinence.^30,35 A recent review concluded that “the magnitude of neuropsychological impairment and the extent to which it persists after abstinence may depend on the frequency and the duration of Cannabis use, length of abstinence, and age at onset of use.”³¹ The effects of HPC on long-term cognitive deficits have not been studied.

Structural brain changes

A number of studies have determined an association between Cannabis use and brain changes involving structures governing memory and emotional processing, including reduced volume of the hippocampus,³⁶ temporal cortex, insula, and orbitofrontal cortex.³⁷ Although many of these changes appear to be dose-related, some morphologic changes have been reported among young recreational users without Cannabis dependence.³⁸ This has resulted in an understandable concern about the effects of Cannabis on the brains of young people with limited exposure; however, it is not yet clear to what extent detected brain changes are pathological and reflect functional deficits.

Recent research using newer neuroimaging modalities provides preliminary support of Cannabis use associated with white matter changes that, in turn, are correlated with cognitive impairment.³⁹ One study comparing low-potency Cannabis and HPC users with and without first-episode psychosis found a significant effect of Cannabis potency on disturbances in white matter microstructural organization in the corpus callosum.⁴⁰ These findings provide sufficient cause for concern that structural brain changes associated with cognitive impairment are more likely to occur with HPC use.

Recommendations for clinicians

Similar to any drug, the effects of THC and its psychiatric sequelae can be expected to increase with dosage. To date, much of the information about psychiatric risks has been based on studies of low- and moderate-potency Cannabis rather than the much higher potency Cannabis products, such as hyper-concentrated “wax dabs,” that are available today. Data from social media suggest that these products may be associated with novel patterns of use, such as with the intention of “passing out.”⁴¹ It is likely that clinicians will encounter greater psychiatric morbidity associated with HPC use.

Although clinicians may be accustomed to asking about the frequency and duration of Cannabis use, it is now prudent also to ask patients about Cannabis potency to better assess the potential risks of use. The potency of different marijuana products is openly advertised within some “medical marijuana” dispensaries, although the accuracy of information in products such as “edibles” has been called into question.⁵

Physicians are increasingly asked to provide recommendations on “medical marijuana” use. A recent paper outlined characteristics of appropriate candidates for “medical marijuana” including:

• having a debilitating condition that might benefit from Cannabis
• multiple failed trials of conventional pharmacotherapies including FDA-approved cannabinoids
• lack of substance use disorders, psychosis, or unstable mood or anxiety disorders
• residence in a state where “medical marijuana” is legal.⁴²

As part of the informed consent process, physicians providing recommendations for “medical marijuana” now must consider the effects of HPC when weighing potential risks against any benefits of Cannabis use. Those monitoring patients using Cannabis should be aware of the potential for greater psychiatric morbidity with HPC and should educate patients about that risk. Failure to adequately warn patients about such morbidity or to screen for risk factors such as psychosis could leave physicians vulnerable to malpractice litigation.

In the United States, the average potency of Cannabis has increased significantly over the past few decades in response to consumer demand and policies in some states that have legalized marijuana for medicinal and recreational purposes.¹ Whereas the delta-9-tetrahydrocannabinol (THC) content of “street” marijuana was <1% in the 1970s and 4% in the 1990s, by 2012, analyses of Cannabis samples seized by law enforcement agencies documented a rise in average THC potency to >12%.^1-3

Although this increase in potency has been overstated in the media because studies did not control for the effects of changes in sampling methods on freshness, it is estimated that Cannabis potency increased 7-fold from 1970 to 2010.³ Also, Cannabis preparations such as hashish and hash oil extracts containing THC well above average—from 35% to 90% THC—are now more widely available. In states where marijuana has been legalized, high-potency Cannabis (HPC) in the form of “edibles” (eg, marijuana added to baked goods, candy, or drinks) and hash oil extracts (Table 1)^4-13 can be readily obtained from dispensaries or even at local farmers’ markets.

The potency of Cannabis, typically defined as the percentage of THC, its chief psychoactive component, varies depending on the genetic strain of the plant, cultivation techniques, and methods of processing and storage. For example, relative to “average marijuana,” hemp (Cannabis bred for industrial purposes) has very little THC, while sinsemilla (flowering buds from unpollinated female plants), hashish (Cannabis resin), and extracted hash oil contain increasing amounts of THC (Table 2).^1,2

Cannabis use disorder

Recent data indicate that the prevalence of Cannabis use disorders (eg, abuse and dependence) in the United States is approximately 3% among the general population and >30% among Cannabis users.¹⁶ The availability of increasingly potent forms of Cannabis has been cited as a possible explanation for this rise, despite no change in the prevalence of overall marijuana use between 1991 to 1992 and 2001 to 2002.¹⁷ However, while the prevalence of marijuana use disorders has continued to rise—nearly doubling from 2001 to 2002 to 2012 to 2013—this latest increase occurred with a significant increase in overall marijuana use, such that the actual rate of Cannabis use disorders among users seems to have plateaued, despite the continued rise in marijuana potency.¹⁶ This discrepancy could be explained if Cannabis users cut back past a specific threshold of increasing potency. However, 2 studies have called into question how effective such titration efforts might be in practice. In one study, Cannabis users who preferred more potent Cannabis inhaled lower volumes of smoke, but did not fully compensate for the increased potency, such that use of HPC still resulted in greater THC exposure.¹⁸ Another study found that HPC users rolled less marijuana into their joints but not enough to mitigate the impact of greater potency.¹⁹ Therefore, it appears that HPC users typically expose themselves to greater amounts of THC, which could place them at higher risk of addiction.

Although a causal association between increasing Cannabis potency and the rate of substance use disorders among users remains unclear based on epidemiologic studies from the United States, a recent study from the United Kingdom examined the impact of Cannabis potency on dependence.²⁰ This cross-sectional survey found that, although HPC was preferred by users and was rated as offering the “best high,” its use was associated with increasing severity of dependence, especially among young people. The limited available evidence supports a greater risk of Cannabis use disorders with increasing potency.

Psychosis

Based on longitudinal studies published over the past 30 years, it is clear that using Cannabis at a young age (age <15 to 18) increases the risk of developing a psychotic disorder.²¹ This association appears to be dose-dependent, with studies consistently demonstrating that psychosis risk increases with greater frequency of Cannabis use.²² The accumulated evidence to date is strong enough to view the psychotic potential of Cannabis as a significant public health concern.²¹

If risk of psychosis is proportional to the amount of Cannabis used as measured by frequency, it follows that this risk might be affected similarly by Cannabis potency. In another paper, I discussed the potential for greater risk of psychosis in the context of medical marijuana and synthetic cannabinoids.²³ My colleagues and I also have published case reports describing emerging psychosis among regular Cannabis users after escalating to higher potency medical marijuana²⁴ and a hyperconcentrated form of hash oil known as Cannabis “wax” or “dabs” that contains as much as 90% THC.⁴ Preliminary anecdotal evidence supports the plausibility of HPC being more psychotoxic than less potent forms.

Several studies from a research group in the United Kingdom, where sinsemilla has increasingly dominated the drug market, likewise have reported that the use of HPC is associated with a greater risk of psychosis. The first of these studies, published in 2009, found that adults hospitalized for first-episode psychosis were more likely to have used HPC than healthy controls.²⁵ Among Cannabis users, HPC use was associated with a 7-fold increased risk of psychosis, with daily HPC use associated with a 12-fold increased risk.

Based on a larger dataset, a second study reported that high-potency, but not low-potency, Cannabis increased the risk of first-episode psychosis with increasing frequency of use.²⁶ Daily users of HPC had a 5-fold higher risk of psychosis compared with those that had never used Cannabis. A third study reported that HPC use and daily Cannabis use were independently associated with an earlier onset of first-episode psychosis, with daily HPC users developing first-episode psychosis an average of 6 years earlier than non-Cannabis users.²⁷ Finally, a prospective study following patients with first-episode psychosis over 2 years found that the greatest risk of relapse—defined by hospital admission caused by exacerbation of psychotic symptoms—was found among self-reported daily users of HPC, while the lowest risk was among those who stopped using Cannabis after their initial psychotic episode.²⁸

The findings from these 4 studies suggest that the increased risk of psychosis with Cannabis is proportional to overall exposure, determined by both frequency of use and Cannabis potency.

Cognition

There is little doubt that using Cannabis can impair cognition acutely, “after all, this is the basic reason for its recreational use,” as one author wrote.²⁹ As with psychosis, the available evidence indicates that the degree of cognitive impairment is related to the frequency and duration of Cannabis use as well as age of onset of use.^30,31

Few studies have assessed cognitive functioning in relation to Cannabis potency with most only examining the effects of relatively low-potency Cannabis with inconsistent results. For example, 2 studies compared cognitive performance in individuals smoking Cannabis with 1.8% and 3.9% THC. One study found that using higher potency Cannabis resulted in prolonged time needed to complete certain cognitive tasks,³² whereas the other found greater impairment in performance on a decision-making task at both potencies compared with non-users but no differences between the 2 dosages.³³ Detecting significant differences may be difficult within the narrow range of low Cannabis potency studied where any findings have limited applicability in the context of today’s Cannabis with much higher THC content.

To date, only 1 study has assessed cognition at higher Cannabis potencies, comparing Cannabis with 4% THC to 13% THC.³⁴ Cognitive impairments increased with higher potency, especially in tasks that measured motor control and executive functioning. Therefore it appears that higher potency Cannabis use is associated with greater acute cognitive impairment.

The longer-term effects on cognition are less clear, with conflicting evidence about whether Cannabis use can result in residual cognitive impairment despite abstinence.^30,35 A recent review concluded that “the magnitude of neuropsychological impairment and the extent to which it persists after abstinence may depend on the frequency and the duration of Cannabis use, length of abstinence, and age at onset of use.”³¹ The effects of HPC on long-term cognitive deficits have not been studied.

Structural brain changes

A number of studies have determined an association between Cannabis use and brain changes involving structures governing memory and emotional processing, including reduced volume of the hippocampus,³⁶ temporal cortex, insula, and orbitofrontal cortex.³⁷ Although many of these changes appear to be dose-related, some morphologic changes have been reported among young recreational users without Cannabis dependence.³⁸ This has resulted in an understandable concern about the effects of Cannabis on the brains of young people with limited exposure; however, it is not yet clear to what extent detected brain changes are pathological and reflect functional deficits.

Recent research using newer neuroimaging modalities provides preliminary support of Cannabis use associated with white matter changes that, in turn, are correlated with cognitive impairment.³⁹ One study comparing low-potency Cannabis and HPC users with and without first-episode psychosis found a significant effect of Cannabis potency on disturbances in white matter microstructural organization in the corpus callosum.⁴⁰ These findings provide sufficient cause for concern that structural brain changes associated with cognitive impairment are more likely to occur with HPC use.

Recommendations for clinicians

Similar to any drug, the effects of THC and its psychiatric sequelae can be expected to increase with dosage. To date, much of the information about psychiatric risks has been based on studies of low- and moderate-potency Cannabis rather than the much higher potency Cannabis products, such as hyper-concentrated “wax dabs,” that are available today. Data from social media suggest that these products may be associated with novel patterns of use, such as with the intention of “passing out.”⁴¹ It is likely that clinicians will encounter greater psychiatric morbidity associated with HPC use.

Although clinicians may be accustomed to asking about the frequency and duration of Cannabis use, it is now prudent also to ask patients about Cannabis potency to better assess the potential risks of use. The potency of different marijuana products is openly advertised within some “medical marijuana” dispensaries, although the accuracy of information in products such as “edibles” has been called into question.⁵

Physicians are increasingly asked to provide recommendations on “medical marijuana” use. A recent paper outlined characteristics of appropriate candidates for “medical marijuana” including:

• having a debilitating condition that might benefit from Cannabis
• multiple failed trials of conventional pharmacotherapies including FDA-approved cannabinoids
• lack of substance use disorders, psychosis, or unstable mood or anxiety disorders
• residence in a state where “medical marijuana” is legal.⁴²

As part of the informed consent process, physicians providing recommendations for “medical marijuana” now must consider the effects of HPC when weighing potential risks against any benefits of Cannabis use. Those monitoring patients using Cannabis should be aware of the potential for greater psychiatric morbidity with HPC and should educate patients about that risk. Failure to adequately warn patients about such morbidity or to screen for risk factors such as psychosis could leave physicians vulnerable to malpractice litigation.

In the United States, the average potency of Cannabis has increased significantly over the past few decades in response to consumer demand and policies in some states that have legalized marijuana for medicinal and recreational purposes.¹ Whereas the delta-9-tetrahydrocannabinol (THC) content of “street” marijuana was <1% in the 1970s and 4% in the 1990s, by 2012, analyses of Cannabis samples seized by law enforcement agencies documented a rise in average THC potency to >12%.^1-3

Although this increase in potency has been overstated in the media because studies did not control for the effects of changes in sampling methods on freshness, it is estimated that Cannabis potency increased 7-fold from 1970 to 2010.³ Also, Cannabis preparations such as hashish and hash oil extracts containing THC well above average—from 35% to 90% THC—are now more widely available. In states where marijuana has been legalized, high-potency Cannabis (HPC) in the form of “edibles” (eg, marijuana added to baked goods, candy, or drinks) and hash oil extracts (Table 1)^4-13 can be readily obtained from dispensaries or even at local farmers’ markets.

The potency of Cannabis, typically defined as the percentage of THC, its chief psychoactive component, varies depending on the genetic strain of the plant, cultivation techniques, and methods of processing and storage. For example, relative to “average marijuana,” hemp (Cannabis bred for industrial purposes) has very little THC, while sinsemilla (flowering buds from unpollinated female plants), hashish (Cannabis resin), and extracted hash oil contain increasing amounts of THC (Table 2).^1,2

Cannabis use disorder

Recent data indicate that the prevalence of Cannabis use disorders (eg, abuse and dependence) in the United States is approximately 3% among the general population and >30% among Cannabis users.¹⁶ The availability of increasingly potent forms of Cannabis has been cited as a possible explanation for this rise, despite no change in the prevalence of overall marijuana use between 1991 to 1992 and 2001 to 2002.¹⁷ However, while the prevalence of marijuana use disorders has continued to rise—nearly doubling from 2001 to 2002 to 2012 to 2013—this latest increase occurred with a significant increase in overall marijuana use, such that the actual rate of Cannabis use disorders among users seems to have plateaued, despite the continued rise in marijuana potency.¹⁶ This discrepancy could be explained if Cannabis users cut back past a specific threshold of increasing potency. However, 2 studies have called into question how effective such titration efforts might be in practice. In one study, Cannabis users who preferred more potent Cannabis inhaled lower volumes of smoke, but did not fully compensate for the increased potency, such that use of HPC still resulted in greater THC exposure.¹⁸ Another study found that HPC users rolled less marijuana into their joints but not enough to mitigate the impact of greater potency.¹⁹ Therefore, it appears that HPC users typically expose themselves to greater amounts of THC, which could place them at higher risk of addiction.

Although a causal association between increasing Cannabis potency and the rate of substance use disorders among users remains unclear based on epidemiologic studies from the United States, a recent study from the United Kingdom examined the impact of Cannabis potency on dependence.²⁰ This cross-sectional survey found that, although HPC was preferred by users and was rated as offering the “best high,” its use was associated with increasing severity of dependence, especially among young people. The limited available evidence supports a greater risk of Cannabis use disorders with increasing potency.

Psychosis

Based on longitudinal studies published over the past 30 years, it is clear that using Cannabis at a young age (age <15 to 18) increases the risk of developing a psychotic disorder.²¹ This association appears to be dose-dependent, with studies consistently demonstrating that psychosis risk increases with greater frequency of Cannabis use.²² The accumulated evidence to date is strong enough to view the psychotic potential of Cannabis as a significant public health concern.²¹

If risk of psychosis is proportional to the amount of Cannabis used as measured by frequency, it follows that this risk might be affected similarly by Cannabis potency. In another paper, I discussed the potential for greater risk of psychosis in the context of medical marijuana and synthetic cannabinoids.²³ My colleagues and I also have published case reports describing emerging psychosis among regular Cannabis users after escalating to higher potency medical marijuana²⁴ and a hyperconcentrated form of hash oil known as Cannabis “wax” or “dabs” that contains as much as 90% THC.⁴ Preliminary anecdotal evidence supports the plausibility of HPC being more psychotoxic than less potent forms.

Several studies from a research group in the United Kingdom, where sinsemilla has increasingly dominated the drug market, likewise have reported that the use of HPC is associated with a greater risk of psychosis. The first of these studies, published in 2009, found that adults hospitalized for first-episode psychosis were more likely to have used HPC than healthy controls.²⁵ Among Cannabis users, HPC use was associated with a 7-fold increased risk of psychosis, with daily HPC use associated with a 12-fold increased risk.

Based on a larger dataset, a second study reported that high-potency, but not low-potency, Cannabis increased the risk of first-episode psychosis with increasing frequency of use.²⁶ Daily users of HPC had a 5-fold higher risk of psychosis compared with those that had never used Cannabis. A third study reported that HPC use and daily Cannabis use were independently associated with an earlier onset of first-episode psychosis, with daily HPC users developing first-episode psychosis an average of 6 years earlier than non-Cannabis users.²⁷ Finally, a prospective study following patients with first-episode psychosis over 2 years found that the greatest risk of relapse—defined by hospital admission caused by exacerbation of psychotic symptoms—was found among self-reported daily users of HPC, while the lowest risk was among those who stopped using Cannabis after their initial psychotic episode.²⁸

The findings from these 4 studies suggest that the increased risk of psychosis with Cannabis is proportional to overall exposure, determined by both frequency of use and Cannabis potency.

Cognition

There is little doubt that using Cannabis can impair cognition acutely, “after all, this is the basic reason for its recreational use,” as one author wrote.²⁹ As with psychosis, the available evidence indicates that the degree of cognitive impairment is related to the frequency and duration of Cannabis use as well as age of onset of use.^30,31

Few studies have assessed cognitive functioning in relation to Cannabis potency with most only examining the effects of relatively low-potency Cannabis with inconsistent results. For example, 2 studies compared cognitive performance in individuals smoking Cannabis with 1.8% and 3.9% THC. One study found that using higher potency Cannabis resulted in prolonged time needed to complete certain cognitive tasks,³² whereas the other found greater impairment in performance on a decision-making task at both potencies compared with non-users but no differences between the 2 dosages.³³ Detecting significant differences may be difficult within the narrow range of low Cannabis potency studied where any findings have limited applicability in the context of today’s Cannabis with much higher THC content.

To date, only 1 study has assessed cognition at higher Cannabis potencies, comparing Cannabis with 4% THC to 13% THC.³⁴ Cognitive impairments increased with higher potency, especially in tasks that measured motor control and executive functioning. Therefore it appears that higher potency Cannabis use is associated with greater acute cognitive impairment.

The longer-term effects on cognition are less clear, with conflicting evidence about whether Cannabis use can result in residual cognitive impairment despite abstinence.^30,35 A recent review concluded that “the magnitude of neuropsychological impairment and the extent to which it persists after abstinence may depend on the frequency and the duration of Cannabis use, length of abstinence, and age at onset of use.”³¹ The effects of HPC on long-term cognitive deficits have not been studied.

Structural brain changes

A number of studies have determined an association between Cannabis use and brain changes involving structures governing memory and emotional processing, including reduced volume of the hippocampus,³⁶ temporal cortex, insula, and orbitofrontal cortex.³⁷ Although many of these changes appear to be dose-related, some morphologic changes have been reported among young recreational users without Cannabis dependence.³⁸ This has resulted in an understandable concern about the effects of Cannabis on the brains of young people with limited exposure; however, it is not yet clear to what extent detected brain changes are pathological and reflect functional deficits.

Recent research using newer neuroimaging modalities provides preliminary support of Cannabis use associated with white matter changes that, in turn, are correlated with cognitive impairment.³⁹ One study comparing low-potency Cannabis and HPC users with and without first-episode psychosis found a significant effect of Cannabis potency on disturbances in white matter microstructural organization in the corpus callosum.⁴⁰ These findings provide sufficient cause for concern that structural brain changes associated with cognitive impairment are more likely to occur with HPC use.

Recommendations for clinicians

Similar to any drug, the effects of THC and its psychiatric sequelae can be expected to increase with dosage. To date, much of the information about psychiatric risks has been based on studies of low- and moderate-potency Cannabis rather than the much higher potency Cannabis products, such as hyper-concentrated “wax dabs,” that are available today. Data from social media suggest that these products may be associated with novel patterns of use, such as with the intention of “passing out.”⁴¹ It is likely that clinicians will encounter greater psychiatric morbidity associated with HPC use.

Although clinicians may be accustomed to asking about the frequency and duration of Cannabis use, it is now prudent also to ask patients about Cannabis potency to better assess the potential risks of use. The potency of different marijuana products is openly advertised within some “medical marijuana” dispensaries, although the accuracy of information in products such as “edibles” has been called into question.⁵

Physicians are increasingly asked to provide recommendations on “medical marijuana” use. A recent paper outlined characteristics of appropriate candidates for “medical marijuana” including:

• having a debilitating condition that might benefit from Cannabis
• multiple failed trials of conventional pharmacotherapies including FDA-approved cannabinoids
• lack of substance use disorders, psychosis, or unstable mood or anxiety disorders
• residence in a state where “medical marijuana” is legal.⁴²

As part of the informed consent process, physicians providing recommendations for “medical marijuana” now must consider the effects of HPC when weighing potential risks against any benefits of Cannabis use. Those monitoring patients using Cannabis should be aware of the potential for greater psychiatric morbidity with HPC and should educate patients about that risk. Failure to adequately warn patients about such morbidity or to screen for risk factors such as psychosis could leave physicians vulnerable to malpractice litigation.