Evidence-Based Reviews

Difficult ethical situations can arise when treating perinatal women who have serious mental illness (SMI). Clinicians must consider ethical issues related to administering antipsychotic medications, the safety of breastfeeding, and concerns for child welfare. They need to carefully weigh the risks and benefits of each decision when treating perinatal women who have SMI. Ethical guidelines can help clinicians best support families in these situations.

In this article, we describe 2 cases of women with psychotic disorders who requested to breastfeed after delivering their child during an inpatient psychiatric hospitalization. The course of their hospitalizations illustrated common ethical questions and facilitated the creation of a framework to assist with complex decision-making regarding breastfeeding on inpatient psychiatric units.

CASE 1

Ms. C, age 41, is multigravida with a psychiatric history of chronic, severe schizoaffective disorder and lives in supportive housing. When Ms. C presents to the hospital in search of a rape kit, clinicians discover she is 22 weeks pregnant but has not received any prenatal care. Psychiatry is consulted because she is found to be intermittently agitated and endorses grandiose delusions. Ms. C requires involuntary hospitalization for decompensated psychosis because she refuses prenatal and psychiatric care. Because it has reassuring reproductive safety data,¹ olanzapine 5 mg/d is started. However, Ms. C experiences minimal improvement from a maximum dose of 20 mg/d. After 13 weeks on the psychiatry unit, she is transferred to obstetrics service for preeclampsia with severe features. Ms. C requires an urgent cesarean delivery at 37 weeks. Her baby boy is transferred to the neonatal intensive care unit (NICU) for transient tachypnea. After delivery and in consultation with psychiatry, the pediatrics team calls Child Protective Services (CPS) due to concern for neglect driven by Ms. C’s psychiatric condition. Ms. C visits the child with medical unit staff supervision in the NICU without consulting with the psychiatry service or CPS. On postpartum Day 2, Ms. C is transferred back to psychiatry for persistent psychosis.

On postpartum Day 3, Ms. C starts to produce breastmilk and requests to breastfeed. At this time, the multidisciplinary team determines she is not able to visit her child in the NICU due to psychiatric instability. No plan is developed to facilitate hand expression or pumping of breastmilk while Ms. C is on the psychiatric unit. The clinical teams discuss whether the benefits of breastfeeding and/or pumping breastmilk would outweigh the risks. CPS determines that Ms. C is unable to retain custody and places the child in kinship foster care while awaiting clinical improvement from her.

CASE 2

Ms. S, age 32, has a history of schizophrenia. She lives with her husband and parents. She is pregnant for the first time and has been receiving consistent prenatal care. Ms. S is brought to the hospital by her husband for bizarre behavior and paranoia after self-discontinuing risperidone 2 mg twice daily due to concern about the medication’s influence on her pregnancy. An ultrasound confirms she is 37 weeks pregnant. Psychiatry is consulted because Ms. S is internally preoccupied, delusional, and endorses auditory hallucinations. She requires involuntary hospitalization for decompensated psychosis. During admission, Ms. S experiences improvement of her psychiatric symptoms while receiving risperidone 2 mg twice daily, which she takes consistently after receiving extensive psychoeducation regarding its safety profile during pregnancy and lactation.

After 2 weeks on the psychiatry unit, Ms. S’s care team transfers her to the obstetrics service with one-to-one supervision. At 39 weeks gestation, she has a vaginal delivery without complications. Because there are no concerns about infant harm, obstetrics, pediatrics, and psychiatry coordinate care so the baby can room in with Ms. S, her husband, and a staff supervisor to facilitate bonding. Ms. S starts to lactate, wishes to breastfeed, and meets with lactation, pediatric, obstetric, and psychiatric specialists to discuss the risks and benefits of breastfeeding and pumping breastmilk. She pursues direct breastfeeding until the baby is discharged home with the husband at postpartum Day 2. CPS is not called because there are no concerns for parental abuse or neglect at the infant’s discharge.

On postpartum Day 2, the obstetrics service transfers Ms. S back to the psychiatric unit for further treatment of her paranoia. She wishes to pump breastmilk while hospitalized, so the treatment team supplies a breast pump, facilitates the storage of breastmilk, and coordinates supervision during pumping to reduce the ligature risk. Ms. S’s husband visits daily to transport the milk and feed the infant breastmilk and formula to meet its nutritional needs. Ms. S maintains psychiatric stability while breast pumping, and the team helps transition her to breastfeeding during visitation with her husband and infant until she is discharged home at 2 weeks postpartum.

Continue to: Approaching care with a relational ethics framework

A relational ethics framework was constructed to evaluate whether to support breastfeeding for both patients during their psychiatric hospitalizations. A relational ethics perspective is defined as “a moral responsibility within a context of human relations” [that] “recognizes the human interdependency and reciprocity within which personal autonomy is embedded.”² This framework values connectedness and commonality between various and even conflicting parties. In the setting of a clinician-patient relationship, health care decisions are made with consideration of the patient’s traditional beliefs, values, and principles rather than the application of impartial moral principles. For these complex cases, this framework was chosen to determine the safest possible outcome for both mother and child.

Risks/benefits of breastfeeding by patients who have SMI

There are several methods of breastfeeding, including direct breastfeeding and other ways of expressing breastmilk such as pumping or hand expression.³ Unlike other forms of feeding using breastmilk, direct breastfeeding has been extensively studied, has well-established medical and psychological benefits for newborns and mothers, and enhances long-term bonding.⁴ Compared with their counterparts who do not breastfeed, mothers who breastfeed have lower rates of unintended pregnancy, cardiovascular disease, postpartum bleeding, osteoporosis, and breast and ovarian cancer.⁵ Among its key psychological benefits, breastfeeding is associated with an increase in maternal self-efficacy and, in some research, has been shown to be associated with a decreased risk of postpartum depression and stress.Additionally, breastfed infants experience lower rates of childhood infection and obesity, and improved nutrition, cognitive development, and immune function.⁶ The American Academy of Pediatrics recognizes these benefits and recommends that women exclusively breastfeed for 6 months postpartum and continue to breastfeed for 2 years or beyond if mutually desired by the mother and child.⁷ Absolute contraindications to breastfeeding must be ruled out (eg, infant classic galactosemia; maternal use of illicit substances such as cocaine, opioids, or phencyclidine; maternal HIV infection, etc).

The risks of breastfeeding by patients who have SMI must also be considered. In severe situations, the infant can be exposed to a mother’s agitation secondary to psychosis.^8,9 The transmission of antipsychotic medication through breastmilk and associated adverse effects (eg, sedation, poor feeding, and extrapyramidal symptoms) are also potential risks and varies among different antipsychotic medications.^1,10 Therefore, when prescribing an antipsychotic for a patient with SMI who breastfeeds, it is crucial to consider the medication’s safety profile as well as other factors, such as the relative infant dose (the weight-adjusted [ie, mg/kg] percentage of the maternal dosage ingested by a fully breastfed infant) and the molecular characteristics of the medication.^10-12 Neonates should be routinely monitored for adverse effects, medication toxicity, and withdrawal symptoms, and care should be coordinated with the infant’s pediatrician. Certain antipsychotic medications, such as aripiprazole, may impact breastmilk production through the dopamine agonist’s interference of the prolactin reflex and anticholinergic properties.^11,13 For a patient with SMI, perhaps the most significant risk involves the time and resources needed for breastfeeding, which can interfere with sleep and psychiatric treatment and possibly further exacerbate psychiatric symptoms.^14-16 Additionally, breastfeeding difficulties or disruption can increase the risk of psychiatric symptoms and psychological distress.¹⁷ In Ms. C’s case, there was a delay in the baby latching as well as multiple medical and psychiatric factors that hindered the milk-ejection reflex to properly initiate; both of these factors rendered breastfeeding particularly difficult while Ms. C was on the inpatient psychiatry unit.¹⁷ In comparison, Ms. S was able to bond with her infant shortly after delivery, which facilitated the milk-ejection reflex and lactation.

Patients who wish to directly breastfeed but struggle to do so while tending to their acute psychiatric condition can benefit from expression of breastmilk that can be provided to the infant or discarded to facilitate breastfeeding in the future.¹⁸ While expression of breastmilk may not be as advantageous for infant health as direct breastfeeding due to the potential changes in breastmilk composition from collecting, storing, and heating, this option can be more protective than formula feeding and facilitate future breastfeeding.¹⁹ In these clinical scenarios, it is standard care to provide a hospital-grade breast pump to the patient, much like a continuous positive airway pressure machine is provided to patients with obstructive sleep apnea.²⁰ However, there is often considerable difficulty obtaining proper breastfeeding equipment and a lack of services devoted to perinatal care in general inpatient settings. Barriers to direct breastfeeding and pumping of breastmilk are highlighted in the Table.²¹

Limitations on breastfeeding on an inpatient unit

The limitations in care and restrictions placed on breastfeeding are more optimally addressed in a mother and baby unit (MBU). MBUs are specialized inpatient psychiatric units designed for mothers experiencing severe perinatal psychiatric difficulties. Unlike general psychiatric units, MBUs allow for joint, full-time admission of mothers and their infants. These units also include multidisciplinary staff who specialize in treating perinatal mental health issues as well as infant care and child development.²² Admission into an MBU is considered best practice for new mothers requiring treatment, particularly in the United Kingdom, Australia, and France, as it is well-recognized that the separation of mother and baby can be psychologically harmful.²³ In the UK, most patients admitted to an MBU showed significant improvement of their psychiatric symptoms and reported overall high satisfaction with care.^24,25 Patients who experience postpartum psychosis prefer MBUs over general psychiatric units because the latter often lack specialized perinatal support, appropriate visitor arrangements, and adequate time with their infant.^26-28

Continue to: The resistance to adopting MBUs in the United States...

The resistance to adopting MBUs in the United States has posed significant barriers in care for perinatal patients and has been attributed to financial barriers, medicolegal risk, staffing, and safety concerns.²⁹ Though currently there are no MBUs in the US, other specialized units have been created. A partial day hospitalization program created in 2000 in Rhode Island for mothers and infants revolutionized the psychiatric care experience for new mothers.³⁰ Since then, other institutions have significantly expanded their services to include perinatal psychiatry inpatient units, yet unlike MBUs, these units typically do not provide overnight rooming-in with infants.³¹ They have the necessary resources and facilities to accommodate the mother’s needs and maximize positive mother-infant interaction, while actively integrating the infant into the mother’s treatment. Breast pumping is treated as a necessary medical procedure and patients can easily access hospital-grade breast pumps with staff supervision. At one such perinatal psychiatric inpatient unit, high rates of treatment satisfaction and significant improvements in symptoms of depression, anxiety, active suicidal ideation, and overall functioning were observed at discharge.³² Therefore, it is crucial to incorporate strategies in general psychiatry units to improve perinatal care, acknowledging that most patients will not have access to these specialized units.²¹

A framework to approaching the relational ethics decisions

An interdisciplinary team used a relational ethics perspective to carefully analyze the risks and benefits of these complex cases. In Figure 1, we propose a framework for the relational ethics decisions of breastfeeding on general inpatient psychiatric units. In creating this framework, we considered principles of autonomy, beneficence, and nonmaleficence, along with the medical and logistical barriers to breastfeeding.

In Ms. C’s case, the team determined that the risks—which included disrupting the mother’s psychiatric treatment, exposing her to psychological harm due to increasing attachment before remanding the child to CPS custody, and risks to the child due to potential unpredictable agitation driven by the treatment-refractory psychosis of the mother as well as that of other psychiatric patients—outweighed the benefits of breastfeeding. We instead recommended breast pumping as an alternative once Ms. C’s psychiatric stability improved. We presented Ms. C with the option of breast pumping on postpartum Day 5. During a 1-day period in which she showed improved behavioral control, she was counseled on the risks and benefits of breastfeeding and exclusive pumping and was notified that the team would help her with the necessary resources, including consultation with a lactation specialist and breast pump. Despite lactation consultant support, Ms. C had low milk production and difficulty with hand expression, which was very discouraging to her. She produced 1 ounce of milk that was shared with the newborn while in the NICU. Because Ms. C’s psychiatric symptoms continued to be severe, with lability and aggression, and because pumping was triggering distress, the multidisciplinary team determined the best course of care would be to focus on her psychiatric recovery rather than on pumping breastmilk. To reduce milk production and minimize discomfort secondary to breast engorgement, the lactation consultant recommended cold compresses, pain management, and compression of breasts. Ultimately, the mother-infant dyad was unable to reap the benefits of breastfeeding (via pumping or direct breastfeeding) due to the mother’s underlying psychiatric illness, although the staffing, psychosocial support, and logistical limitations contributed to this outcome.³³

In Ms. S’s case, the treatment team determined that there were no medical or psychiatric contraindications to breastfeeding, and she was counseled on the risks and benefits of direct breastfeeding and pumping. The treatment team determined it was safe for Ms. S to directly breastfeed as there were no concerns for infant harm post­delivery with constant supervision while on the obstetrics floor. The patient opted to directly breastfeed, which was successful with the guidance of a lactation specialist. When she was transferred to the psychiatric unit on postpartum Day 2, her child was discharged home with the husband. The patient was then encouraged to pump while the psychiatrists monitored her symptoms closely and facilitated increased staff and resources. Transportation of breastmilk was made possible by the family, and on postpartum Day 5, as the patient maintained psychiatric stability, the team discussed with Ms. S and her husband the prospect of direct breastfeeding. The treatment team arranged for separate visitation hours to minimize the possibility of exposing the infant to aggression from other patients on the unit and advocated with hospital leadership to approve of infant visitation on the unit.

Impact of involvement of Child Protective Services

The involvement of CPS also added complexity to Ms. C’s case. Without proper legal guidance, mothers with psychosis who lose custody can find it difficult to navigate the legal system and maintain contact with their children.³⁴ As the prevalence of custody loss in mothers with psychosis is high (approximately 50% according to research published in the last 10 years), effective interventions to reunite the mother and child must be promoted (Figure 2).^35-39 Ultimately, the goal of psychiatric hospitalization for perinatal women who have SMI is psychiatric stabilization. The preemptive involvement of psychiatry is crucial because it can allow for early postpartum planning and can provide an opportunity to address feeding options and custody concerns with the patient, social supports and services, and various medical teams. In Ms. C’s case, she visited her baby in the NICU on postpartum Day 2 without consultation with psychiatry or CPS, which posed risks to the patient, infant, and staff. It is vital that various clinicians collaborate with each other and the patient, working towards the goal of optimizing the patient’s mental health to allow for parenting rights in the future and maximizing a sustainable attachment between the parent and child. In Ms. S’s case, the husband was able to facilitate caring for the baby while the mother was hospitalized and played an integral role in the feeding process via pumped breastmilk and transport of the infant for direct breastfeeding.

Continue to: The differences in these 2 cases...

The differences in these 2 cases show the extreme importance of social support to benefit both the mother and child, and the need for more comprehensive social services for women who do not have a social safety net.

Bottom Line

These complex cases highlight an ethical decision-making approach to breastfeeding in perinatal women who have serious mental illness. Collaborative care and shared decision-making, which highlight the interests of the mother and baby, are crucial when assessing the risks and benefits of breastfeeding and pumping breastmilk. Our relational ethics framework can be used to better evaluate and implement breastfeeding options on general psychiatric units.

Related Resources

• Tillman B, Sloan N, Westmoreland P. How COVID-19 affects peripartum women’s mental health. Current Psychiatry. 2021;20(6):18-22. doi:10.12788/cp.0129
• Koch J, Preinitz J. Antidepressants for patients who are breastfeeding: what to consider. Current Psychiatry. 2023;22(5):20-23,48. doi:10.12788/cp.0355

Drug Brand Names

Aripiprazole • Abilify
Olanzapine • Zyprexa
Risperidone • Risperdal

Difficult ethical situations can arise when treating perinatal women who have serious mental illness (SMI). Clinicians must consider ethical issues related to administering antipsychotic medications, the safety of breastfeeding, and concerns for child welfare. They need to carefully weigh the risks and benefits of each decision when treating perinatal women who have SMI. Ethical guidelines can help clinicians best support families in these situations.

In this article, we describe 2 cases of women with psychotic disorders who requested to breastfeed after delivering their child during an inpatient psychiatric hospitalization. The course of their hospitalizations illustrated common ethical questions and facilitated the creation of a framework to assist with complex decision-making regarding breastfeeding on inpatient psychiatric units.

CASE 1

Ms. C, age 41, is multigravida with a psychiatric history of chronic, severe schizoaffective disorder and lives in supportive housing. When Ms. C presents to the hospital in search of a rape kit, clinicians discover she is 22 weeks pregnant but has not received any prenatal care. Psychiatry is consulted because she is found to be intermittently agitated and endorses grandiose delusions. Ms. C requires involuntary hospitalization for decompensated psychosis because she refuses prenatal and psychiatric care. Because it has reassuring reproductive safety data,¹ olanzapine 5 mg/d is started. However, Ms. C experiences minimal improvement from a maximum dose of 20 mg/d. After 13 weeks on the psychiatry unit, she is transferred to obstetrics service for preeclampsia with severe features. Ms. C requires an urgent cesarean delivery at 37 weeks. Her baby boy is transferred to the neonatal intensive care unit (NICU) for transient tachypnea. After delivery and in consultation with psychiatry, the pediatrics team calls Child Protective Services (CPS) due to concern for neglect driven by Ms. C’s psychiatric condition. Ms. C visits the child with medical unit staff supervision in the NICU without consulting with the psychiatry service or CPS. On postpartum Day 2, Ms. C is transferred back to psychiatry for persistent psychosis.

On postpartum Day 3, Ms. C starts to produce breastmilk and requests to breastfeed. At this time, the multidisciplinary team determines she is not able to visit her child in the NICU due to psychiatric instability. No plan is developed to facilitate hand expression or pumping of breastmilk while Ms. C is on the psychiatric unit. The clinical teams discuss whether the benefits of breastfeeding and/or pumping breastmilk would outweigh the risks. CPS determines that Ms. C is unable to retain custody and places the child in kinship foster care while awaiting clinical improvement from her.

CASE 2

Ms. S, age 32, has a history of schizophrenia. She lives with her husband and parents. She is pregnant for the first time and has been receiving consistent prenatal care. Ms. S is brought to the hospital by her husband for bizarre behavior and paranoia after self-discontinuing risperidone 2 mg twice daily due to concern about the medication’s influence on her pregnancy. An ultrasound confirms she is 37 weeks pregnant. Psychiatry is consulted because Ms. S is internally preoccupied, delusional, and endorses auditory hallucinations. She requires involuntary hospitalization for decompensated psychosis. During admission, Ms. S experiences improvement of her psychiatric symptoms while receiving risperidone 2 mg twice daily, which she takes consistently after receiving extensive psychoeducation regarding its safety profile during pregnancy and lactation.

After 2 weeks on the psychiatry unit, Ms. S’s care team transfers her to the obstetrics service with one-to-one supervision. At 39 weeks gestation, she has a vaginal delivery without complications. Because there are no concerns about infant harm, obstetrics, pediatrics, and psychiatry coordinate care so the baby can room in with Ms. S, her husband, and a staff supervisor to facilitate bonding. Ms. S starts to lactate, wishes to breastfeed, and meets with lactation, pediatric, obstetric, and psychiatric specialists to discuss the risks and benefits of breastfeeding and pumping breastmilk. She pursues direct breastfeeding until the baby is discharged home with the husband at postpartum Day 2. CPS is not called because there are no concerns for parental abuse or neglect at the infant’s discharge.

On postpartum Day 2, the obstetrics service transfers Ms. S back to the psychiatric unit for further treatment of her paranoia. She wishes to pump breastmilk while hospitalized, so the treatment team supplies a breast pump, facilitates the storage of breastmilk, and coordinates supervision during pumping to reduce the ligature risk. Ms. S’s husband visits daily to transport the milk and feed the infant breastmilk and formula to meet its nutritional needs. Ms. S maintains psychiatric stability while breast pumping, and the team helps transition her to breastfeeding during visitation with her husband and infant until she is discharged home at 2 weeks postpartum.

Continue to: Approaching care with a relational ethics framework

A relational ethics framework was constructed to evaluate whether to support breastfeeding for both patients during their psychiatric hospitalizations. A relational ethics perspective is defined as “a moral responsibility within a context of human relations” [that] “recognizes the human interdependency and reciprocity within which personal autonomy is embedded.”² This framework values connectedness and commonality between various and even conflicting parties. In the setting of a clinician-patient relationship, health care decisions are made with consideration of the patient’s traditional beliefs, values, and principles rather than the application of impartial moral principles. For these complex cases, this framework was chosen to determine the safest possible outcome for both mother and child.

Risks/benefits of breastfeeding by patients who have SMI

There are several methods of breastfeeding, including direct breastfeeding and other ways of expressing breastmilk such as pumping or hand expression.³ Unlike other forms of feeding using breastmilk, direct breastfeeding has been extensively studied, has well-established medical and psychological benefits for newborns and mothers, and enhances long-term bonding.⁴ Compared with their counterparts who do not breastfeed, mothers who breastfeed have lower rates of unintended pregnancy, cardiovascular disease, postpartum bleeding, osteoporosis, and breast and ovarian cancer.⁵ Among its key psychological benefits, breastfeeding is associated with an increase in maternal self-efficacy and, in some research, has been shown to be associated with a decreased risk of postpartum depression and stress.Additionally, breastfed infants experience lower rates of childhood infection and obesity, and improved nutrition, cognitive development, and immune function.⁶ The American Academy of Pediatrics recognizes these benefits and recommends that women exclusively breastfeed for 6 months postpartum and continue to breastfeed for 2 years or beyond if mutually desired by the mother and child.⁷ Absolute contraindications to breastfeeding must be ruled out (eg, infant classic galactosemia; maternal use of illicit substances such as cocaine, opioids, or phencyclidine; maternal HIV infection, etc).

The risks of breastfeeding by patients who have SMI must also be considered. In severe situations, the infant can be exposed to a mother’s agitation secondary to psychosis.^8,9 The transmission of antipsychotic medication through breastmilk and associated adverse effects (eg, sedation, poor feeding, and extrapyramidal symptoms) are also potential risks and varies among different antipsychotic medications.^1,10 Therefore, when prescribing an antipsychotic for a patient with SMI who breastfeeds, it is crucial to consider the medication’s safety profile as well as other factors, such as the relative infant dose (the weight-adjusted [ie, mg/kg] percentage of the maternal dosage ingested by a fully breastfed infant) and the molecular characteristics of the medication.^10-12 Neonates should be routinely monitored for adverse effects, medication toxicity, and withdrawal symptoms, and care should be coordinated with the infant’s pediatrician. Certain antipsychotic medications, such as aripiprazole, may impact breastmilk production through the dopamine agonist’s interference of the prolactin reflex and anticholinergic properties.^11,13 For a patient with SMI, perhaps the most significant risk involves the time and resources needed for breastfeeding, which can interfere with sleep and psychiatric treatment and possibly further exacerbate psychiatric symptoms.^14-16 Additionally, breastfeeding difficulties or disruption can increase the risk of psychiatric symptoms and psychological distress.¹⁷ In Ms. C’s case, there was a delay in the baby latching as well as multiple medical and psychiatric factors that hindered the milk-ejection reflex to properly initiate; both of these factors rendered breastfeeding particularly difficult while Ms. C was on the inpatient psychiatry unit.¹⁷ In comparison, Ms. S was able to bond with her infant shortly after delivery, which facilitated the milk-ejection reflex and lactation.

Patients who wish to directly breastfeed but struggle to do so while tending to their acute psychiatric condition can benefit from expression of breastmilk that can be provided to the infant or discarded to facilitate breastfeeding in the future.¹⁸ While expression of breastmilk may not be as advantageous for infant health as direct breastfeeding due to the potential changes in breastmilk composition from collecting, storing, and heating, this option can be more protective than formula feeding and facilitate future breastfeeding.¹⁹ In these clinical scenarios, it is standard care to provide a hospital-grade breast pump to the patient, much like a continuous positive airway pressure machine is provided to patients with obstructive sleep apnea.²⁰ However, there is often considerable difficulty obtaining proper breastfeeding equipment and a lack of services devoted to perinatal care in general inpatient settings. Barriers to direct breastfeeding and pumping of breastmilk are highlighted in the Table.²¹

Limitations on breastfeeding on an inpatient unit

The limitations in care and restrictions placed on breastfeeding are more optimally addressed in a mother and baby unit (MBU). MBUs are specialized inpatient psychiatric units designed for mothers experiencing severe perinatal psychiatric difficulties. Unlike general psychiatric units, MBUs allow for joint, full-time admission of mothers and their infants. These units also include multidisciplinary staff who specialize in treating perinatal mental health issues as well as infant care and child development.²² Admission into an MBU is considered best practice for new mothers requiring treatment, particularly in the United Kingdom, Australia, and France, as it is well-recognized that the separation of mother and baby can be psychologically harmful.²³ In the UK, most patients admitted to an MBU showed significant improvement of their psychiatric symptoms and reported overall high satisfaction with care.^24,25 Patients who experience postpartum psychosis prefer MBUs over general psychiatric units because the latter often lack specialized perinatal support, appropriate visitor arrangements, and adequate time with their infant.^26-28

Continue to: The resistance to adopting MBUs in the United States...

The resistance to adopting MBUs in the United States has posed significant barriers in care for perinatal patients and has been attributed to financial barriers, medicolegal risk, staffing, and safety concerns.²⁹ Though currently there are no MBUs in the US, other specialized units have been created. A partial day hospitalization program created in 2000 in Rhode Island for mothers and infants revolutionized the psychiatric care experience for new mothers.³⁰ Since then, other institutions have significantly expanded their services to include perinatal psychiatry inpatient units, yet unlike MBUs, these units typically do not provide overnight rooming-in with infants.³¹ They have the necessary resources and facilities to accommodate the mother’s needs and maximize positive mother-infant interaction, while actively integrating the infant into the mother’s treatment. Breast pumping is treated as a necessary medical procedure and patients can easily access hospital-grade breast pumps with staff supervision. At one such perinatal psychiatric inpatient unit, high rates of treatment satisfaction and significant improvements in symptoms of depression, anxiety, active suicidal ideation, and overall functioning were observed at discharge.³² Therefore, it is crucial to incorporate strategies in general psychiatry units to improve perinatal care, acknowledging that most patients will not have access to these specialized units.²¹

A framework to approaching the relational ethics decisions

An interdisciplinary team used a relational ethics perspective to carefully analyze the risks and benefits of these complex cases. In Figure 1, we propose a framework for the relational ethics decisions of breastfeeding on general inpatient psychiatric units. In creating this framework, we considered principles of autonomy, beneficence, and nonmaleficence, along with the medical and logistical barriers to breastfeeding.

In Ms. C’s case, the team determined that the risks—which included disrupting the mother’s psychiatric treatment, exposing her to psychological harm due to increasing attachment before remanding the child to CPS custody, and risks to the child due to potential unpredictable agitation driven by the treatment-refractory psychosis of the mother as well as that of other psychiatric patients—outweighed the benefits of breastfeeding. We instead recommended breast pumping as an alternative once Ms. C’s psychiatric stability improved. We presented Ms. C with the option of breast pumping on postpartum Day 5. During a 1-day period in which she showed improved behavioral control, she was counseled on the risks and benefits of breastfeeding and exclusive pumping and was notified that the team would help her with the necessary resources, including consultation with a lactation specialist and breast pump. Despite lactation consultant support, Ms. C had low milk production and difficulty with hand expression, which was very discouraging to her. She produced 1 ounce of milk that was shared with the newborn while in the NICU. Because Ms. C’s psychiatric symptoms continued to be severe, with lability and aggression, and because pumping was triggering distress, the multidisciplinary team determined the best course of care would be to focus on her psychiatric recovery rather than on pumping breastmilk. To reduce milk production and minimize discomfort secondary to breast engorgement, the lactation consultant recommended cold compresses, pain management, and compression of breasts. Ultimately, the mother-infant dyad was unable to reap the benefits of breastfeeding (via pumping or direct breastfeeding) due to the mother’s underlying psychiatric illness, although the staffing, psychosocial support, and logistical limitations contributed to this outcome.³³

In Ms. S’s case, the treatment team determined that there were no medical or psychiatric contraindications to breastfeeding, and she was counseled on the risks and benefits of direct breastfeeding and pumping. The treatment team determined it was safe for Ms. S to directly breastfeed as there were no concerns for infant harm post­delivery with constant supervision while on the obstetrics floor. The patient opted to directly breastfeed, which was successful with the guidance of a lactation specialist. When she was transferred to the psychiatric unit on postpartum Day 2, her child was discharged home with the husband. The patient was then encouraged to pump while the psychiatrists monitored her symptoms closely and facilitated increased staff and resources. Transportation of breastmilk was made possible by the family, and on postpartum Day 5, as the patient maintained psychiatric stability, the team discussed with Ms. S and her husband the prospect of direct breastfeeding. The treatment team arranged for separate visitation hours to minimize the possibility of exposing the infant to aggression from other patients on the unit and advocated with hospital leadership to approve of infant visitation on the unit.

Impact of involvement of Child Protective Services

The involvement of CPS also added complexity to Ms. C’s case. Without proper legal guidance, mothers with psychosis who lose custody can find it difficult to navigate the legal system and maintain contact with their children.³⁴ As the prevalence of custody loss in mothers with psychosis is high (approximately 50% according to research published in the last 10 years), effective interventions to reunite the mother and child must be promoted (Figure 2).^35-39 Ultimately, the goal of psychiatric hospitalization for perinatal women who have SMI is psychiatric stabilization. The preemptive involvement of psychiatry is crucial because it can allow for early postpartum planning and can provide an opportunity to address feeding options and custody concerns with the patient, social supports and services, and various medical teams. In Ms. C’s case, she visited her baby in the NICU on postpartum Day 2 without consultation with psychiatry or CPS, which posed risks to the patient, infant, and staff. It is vital that various clinicians collaborate with each other and the patient, working towards the goal of optimizing the patient’s mental health to allow for parenting rights in the future and maximizing a sustainable attachment between the parent and child. In Ms. S’s case, the husband was able to facilitate caring for the baby while the mother was hospitalized and played an integral role in the feeding process via pumped breastmilk and transport of the infant for direct breastfeeding.

Continue to: The differences in these 2 cases...

The differences in these 2 cases show the extreme importance of social support to benefit both the mother and child, and the need for more comprehensive social services for women who do not have a social safety net.

Bottom Line

These complex cases highlight an ethical decision-making approach to breastfeeding in perinatal women who have serious mental illness. Collaborative care and shared decision-making, which highlight the interests of the mother and baby, are crucial when assessing the risks and benefits of breastfeeding and pumping breastmilk. Our relational ethics framework can be used to better evaluate and implement breastfeeding options on general psychiatric units.

Related Resources

• Tillman B, Sloan N, Westmoreland P. How COVID-19 affects peripartum women’s mental health. Current Psychiatry. 2021;20(6):18-22. doi:10.12788/cp.0129
• Koch J, Preinitz J. Antidepressants for patients who are breastfeeding: what to consider. Current Psychiatry. 2023;22(5):20-23,48. doi:10.12788/cp.0355

Drug Brand Names

Aripiprazole • Abilify
Olanzapine • Zyprexa
Risperidone • Risperdal

Difficult ethical situations can arise when treating perinatal women who have serious mental illness (SMI). Clinicians must consider ethical issues related to administering antipsychotic medications, the safety of breastfeeding, and concerns for child welfare. They need to carefully weigh the risks and benefits of each decision when treating perinatal women who have SMI. Ethical guidelines can help clinicians best support families in these situations.

In this article, we describe 2 cases of women with psychotic disorders who requested to breastfeed after delivering their child during an inpatient psychiatric hospitalization. The course of their hospitalizations illustrated common ethical questions and facilitated the creation of a framework to assist with complex decision-making regarding breastfeeding on inpatient psychiatric units.

CASE 1

Ms. C, age 41, is multigravida with a psychiatric history of chronic, severe schizoaffective disorder and lives in supportive housing. When Ms. C presents to the hospital in search of a rape kit, clinicians discover she is 22 weeks pregnant but has not received any prenatal care. Psychiatry is consulted because she is found to be intermittently agitated and endorses grandiose delusions. Ms. C requires involuntary hospitalization for decompensated psychosis because she refuses prenatal and psychiatric care. Because it has reassuring reproductive safety data,¹ olanzapine 5 mg/d is started. However, Ms. C experiences minimal improvement from a maximum dose of 20 mg/d. After 13 weeks on the psychiatry unit, she is transferred to obstetrics service for preeclampsia with severe features. Ms. C requires an urgent cesarean delivery at 37 weeks. Her baby boy is transferred to the neonatal intensive care unit (NICU) for transient tachypnea. After delivery and in consultation with psychiatry, the pediatrics team calls Child Protective Services (CPS) due to concern for neglect driven by Ms. C’s psychiatric condition. Ms. C visits the child with medical unit staff supervision in the NICU without consulting with the psychiatry service or CPS. On postpartum Day 2, Ms. C is transferred back to psychiatry for persistent psychosis.

On postpartum Day 3, Ms. C starts to produce breastmilk and requests to breastfeed. At this time, the multidisciplinary team determines she is not able to visit her child in the NICU due to psychiatric instability. No plan is developed to facilitate hand expression or pumping of breastmilk while Ms. C is on the psychiatric unit. The clinical teams discuss whether the benefits of breastfeeding and/or pumping breastmilk would outweigh the risks. CPS determines that Ms. C is unable to retain custody and places the child in kinship foster care while awaiting clinical improvement from her.

CASE 2

Ms. S, age 32, has a history of schizophrenia. She lives with her husband and parents. She is pregnant for the first time and has been receiving consistent prenatal care. Ms. S is brought to the hospital by her husband for bizarre behavior and paranoia after self-discontinuing risperidone 2 mg twice daily due to concern about the medication’s influence on her pregnancy. An ultrasound confirms she is 37 weeks pregnant. Psychiatry is consulted because Ms. S is internally preoccupied, delusional, and endorses auditory hallucinations. She requires involuntary hospitalization for decompensated psychosis. During admission, Ms. S experiences improvement of her psychiatric symptoms while receiving risperidone 2 mg twice daily, which she takes consistently after receiving extensive psychoeducation regarding its safety profile during pregnancy and lactation.

After 2 weeks on the psychiatry unit, Ms. S’s care team transfers her to the obstetrics service with one-to-one supervision. At 39 weeks gestation, she has a vaginal delivery without complications. Because there are no concerns about infant harm, obstetrics, pediatrics, and psychiatry coordinate care so the baby can room in with Ms. S, her husband, and a staff supervisor to facilitate bonding. Ms. S starts to lactate, wishes to breastfeed, and meets with lactation, pediatric, obstetric, and psychiatric specialists to discuss the risks and benefits of breastfeeding and pumping breastmilk. She pursues direct breastfeeding until the baby is discharged home with the husband at postpartum Day 2. CPS is not called because there are no concerns for parental abuse or neglect at the infant’s discharge.

On postpartum Day 2, the obstetrics service transfers Ms. S back to the psychiatric unit for further treatment of her paranoia. She wishes to pump breastmilk while hospitalized, so the treatment team supplies a breast pump, facilitates the storage of breastmilk, and coordinates supervision during pumping to reduce the ligature risk. Ms. S’s husband visits daily to transport the milk and feed the infant breastmilk and formula to meet its nutritional needs. Ms. S maintains psychiatric stability while breast pumping, and the team helps transition her to breastfeeding during visitation with her husband and infant until she is discharged home at 2 weeks postpartum.

Continue to: Approaching care with a relational ethics framework

A relational ethics framework was constructed to evaluate whether to support breastfeeding for both patients during their psychiatric hospitalizations. A relational ethics perspective is defined as “a moral responsibility within a context of human relations” [that] “recognizes the human interdependency and reciprocity within which personal autonomy is embedded.”² This framework values connectedness and commonality between various and even conflicting parties. In the setting of a clinician-patient relationship, health care decisions are made with consideration of the patient’s traditional beliefs, values, and principles rather than the application of impartial moral principles. For these complex cases, this framework was chosen to determine the safest possible outcome for both mother and child.

Risks/benefits of breastfeeding by patients who have SMI

There are several methods of breastfeeding, including direct breastfeeding and other ways of expressing breastmilk such as pumping or hand expression.³ Unlike other forms of feeding using breastmilk, direct breastfeeding has been extensively studied, has well-established medical and psychological benefits for newborns and mothers, and enhances long-term bonding.⁴ Compared with their counterparts who do not breastfeed, mothers who breastfeed have lower rates of unintended pregnancy, cardiovascular disease, postpartum bleeding, osteoporosis, and breast and ovarian cancer.⁵ Among its key psychological benefits, breastfeeding is associated with an increase in maternal self-efficacy and, in some research, has been shown to be associated with a decreased risk of postpartum depression and stress.Additionally, breastfed infants experience lower rates of childhood infection and obesity, and improved nutrition, cognitive development, and immune function.⁶ The American Academy of Pediatrics recognizes these benefits and recommends that women exclusively breastfeed for 6 months postpartum and continue to breastfeed for 2 years or beyond if mutually desired by the mother and child.⁷ Absolute contraindications to breastfeeding must be ruled out (eg, infant classic galactosemia; maternal use of illicit substances such as cocaine, opioids, or phencyclidine; maternal HIV infection, etc).

The risks of breastfeeding by patients who have SMI must also be considered. In severe situations, the infant can be exposed to a mother’s agitation secondary to psychosis.^8,9 The transmission of antipsychotic medication through breastmilk and associated adverse effects (eg, sedation, poor feeding, and extrapyramidal symptoms) are also potential risks and varies among different antipsychotic medications.^1,10 Therefore, when prescribing an antipsychotic for a patient with SMI who breastfeeds, it is crucial to consider the medication’s safety profile as well as other factors, such as the relative infant dose (the weight-adjusted [ie, mg/kg] percentage of the maternal dosage ingested by a fully breastfed infant) and the molecular characteristics of the medication.^10-12 Neonates should be routinely monitored for adverse effects, medication toxicity, and withdrawal symptoms, and care should be coordinated with the infant’s pediatrician. Certain antipsychotic medications, such as aripiprazole, may impact breastmilk production through the dopamine agonist’s interference of the prolactin reflex and anticholinergic properties.^11,13 For a patient with SMI, perhaps the most significant risk involves the time and resources needed for breastfeeding, which can interfere with sleep and psychiatric treatment and possibly further exacerbate psychiatric symptoms.^14-16 Additionally, breastfeeding difficulties or disruption can increase the risk of psychiatric symptoms and psychological distress.¹⁷ In Ms. C’s case, there was a delay in the baby latching as well as multiple medical and psychiatric factors that hindered the milk-ejection reflex to properly initiate; both of these factors rendered breastfeeding particularly difficult while Ms. C was on the inpatient psychiatry unit.¹⁷ In comparison, Ms. S was able to bond with her infant shortly after delivery, which facilitated the milk-ejection reflex and lactation.

Patients who wish to directly breastfeed but struggle to do so while tending to their acute psychiatric condition can benefit from expression of breastmilk that can be provided to the infant or discarded to facilitate breastfeeding in the future.¹⁸ While expression of breastmilk may not be as advantageous for infant health as direct breastfeeding due to the potential changes in breastmilk composition from collecting, storing, and heating, this option can be more protective than formula feeding and facilitate future breastfeeding.¹⁹ In these clinical scenarios, it is standard care to provide a hospital-grade breast pump to the patient, much like a continuous positive airway pressure machine is provided to patients with obstructive sleep apnea.²⁰ However, there is often considerable difficulty obtaining proper breastfeeding equipment and a lack of services devoted to perinatal care in general inpatient settings. Barriers to direct breastfeeding and pumping of breastmilk are highlighted in the Table.²¹

Limitations on breastfeeding on an inpatient unit

The limitations in care and restrictions placed on breastfeeding are more optimally addressed in a mother and baby unit (MBU). MBUs are specialized inpatient psychiatric units designed for mothers experiencing severe perinatal psychiatric difficulties. Unlike general psychiatric units, MBUs allow for joint, full-time admission of mothers and their infants. These units also include multidisciplinary staff who specialize in treating perinatal mental health issues as well as infant care and child development.²² Admission into an MBU is considered best practice for new mothers requiring treatment, particularly in the United Kingdom, Australia, and France, as it is well-recognized that the separation of mother and baby can be psychologically harmful.²³ In the UK, most patients admitted to an MBU showed significant improvement of their psychiatric symptoms and reported overall high satisfaction with care.^24,25 Patients who experience postpartum psychosis prefer MBUs over general psychiatric units because the latter often lack specialized perinatal support, appropriate visitor arrangements, and adequate time with their infant.^26-28

Continue to: The resistance to adopting MBUs in the United States...

The resistance to adopting MBUs in the United States has posed significant barriers in care for perinatal patients and has been attributed to financial barriers, medicolegal risk, staffing, and safety concerns.²⁹ Though currently there are no MBUs in the US, other specialized units have been created. A partial day hospitalization program created in 2000 in Rhode Island for mothers and infants revolutionized the psychiatric care experience for new mothers.³⁰ Since then, other institutions have significantly expanded their services to include perinatal psychiatry inpatient units, yet unlike MBUs, these units typically do not provide overnight rooming-in with infants.³¹ They have the necessary resources and facilities to accommodate the mother’s needs and maximize positive mother-infant interaction, while actively integrating the infant into the mother’s treatment. Breast pumping is treated as a necessary medical procedure and patients can easily access hospital-grade breast pumps with staff supervision. At one such perinatal psychiatric inpatient unit, high rates of treatment satisfaction and significant improvements in symptoms of depression, anxiety, active suicidal ideation, and overall functioning were observed at discharge.³² Therefore, it is crucial to incorporate strategies in general psychiatry units to improve perinatal care, acknowledging that most patients will not have access to these specialized units.²¹

A framework to approaching the relational ethics decisions

An interdisciplinary team used a relational ethics perspective to carefully analyze the risks and benefits of these complex cases. In Figure 1, we propose a framework for the relational ethics decisions of breastfeeding on general inpatient psychiatric units. In creating this framework, we considered principles of autonomy, beneficence, and nonmaleficence, along with the medical and logistical barriers to breastfeeding.

In Ms. C’s case, the team determined that the risks—which included disrupting the mother’s psychiatric treatment, exposing her to psychological harm due to increasing attachment before remanding the child to CPS custody, and risks to the child due to potential unpredictable agitation driven by the treatment-refractory psychosis of the mother as well as that of other psychiatric patients—outweighed the benefits of breastfeeding. We instead recommended breast pumping as an alternative once Ms. C’s psychiatric stability improved. We presented Ms. C with the option of breast pumping on postpartum Day 5. During a 1-day period in which she showed improved behavioral control, she was counseled on the risks and benefits of breastfeeding and exclusive pumping and was notified that the team would help her with the necessary resources, including consultation with a lactation specialist and breast pump. Despite lactation consultant support, Ms. C had low milk production and difficulty with hand expression, which was very discouraging to her. She produced 1 ounce of milk that was shared with the newborn while in the NICU. Because Ms. C’s psychiatric symptoms continued to be severe, with lability and aggression, and because pumping was triggering distress, the multidisciplinary team determined the best course of care would be to focus on her psychiatric recovery rather than on pumping breastmilk. To reduce milk production and minimize discomfort secondary to breast engorgement, the lactation consultant recommended cold compresses, pain management, and compression of breasts. Ultimately, the mother-infant dyad was unable to reap the benefits of breastfeeding (via pumping or direct breastfeeding) due to the mother’s underlying psychiatric illness, although the staffing, psychosocial support, and logistical limitations contributed to this outcome.³³

In Ms. S’s case, the treatment team determined that there were no medical or psychiatric contraindications to breastfeeding, and she was counseled on the risks and benefits of direct breastfeeding and pumping. The treatment team determined it was safe for Ms. S to directly breastfeed as there were no concerns for infant harm post­delivery with constant supervision while on the obstetrics floor. The patient opted to directly breastfeed, which was successful with the guidance of a lactation specialist. When she was transferred to the psychiatric unit on postpartum Day 2, her child was discharged home with the husband. The patient was then encouraged to pump while the psychiatrists monitored her symptoms closely and facilitated increased staff and resources. Transportation of breastmilk was made possible by the family, and on postpartum Day 5, as the patient maintained psychiatric stability, the team discussed with Ms. S and her husband the prospect of direct breastfeeding. The treatment team arranged for separate visitation hours to minimize the possibility of exposing the infant to aggression from other patients on the unit and advocated with hospital leadership to approve of infant visitation on the unit.

Impact of involvement of Child Protective Services

The involvement of CPS also added complexity to Ms. C’s case. Without proper legal guidance, mothers with psychosis who lose custody can find it difficult to navigate the legal system and maintain contact with their children.³⁴ As the prevalence of custody loss in mothers with psychosis is high (approximately 50% according to research published in the last 10 years), effective interventions to reunite the mother and child must be promoted (Figure 2).^35-39 Ultimately, the goal of psychiatric hospitalization for perinatal women who have SMI is psychiatric stabilization. The preemptive involvement of psychiatry is crucial because it can allow for early postpartum planning and can provide an opportunity to address feeding options and custody concerns with the patient, social supports and services, and various medical teams. In Ms. C’s case, she visited her baby in the NICU on postpartum Day 2 without consultation with psychiatry or CPS, which posed risks to the patient, infant, and staff. It is vital that various clinicians collaborate with each other and the patient, working towards the goal of optimizing the patient’s mental health to allow for parenting rights in the future and maximizing a sustainable attachment between the parent and child. In Ms. S’s case, the husband was able to facilitate caring for the baby while the mother was hospitalized and played an integral role in the feeding process via pumped breastmilk and transport of the infant for direct breastfeeding.

Continue to: The differences in these 2 cases...

The differences in these 2 cases show the extreme importance of social support to benefit both the mother and child, and the need for more comprehensive social services for women who do not have a social safety net.

Bottom Line

These complex cases highlight an ethical decision-making approach to breastfeeding in perinatal women who have serious mental illness. Collaborative care and shared decision-making, which highlight the interests of the mother and baby, are crucial when assessing the risks and benefits of breastfeeding and pumping breastmilk. Our relational ethics framework can be used to better evaluate and implement breastfeeding options on general psychiatric units.

Related Resources

• Tillman B, Sloan N, Westmoreland P. How COVID-19 affects peripartum women’s mental health. Current Psychiatry. 2021;20(6):18-22. doi:10.12788/cp.0129
• Koch J, Preinitz J. Antidepressants for patients who are breastfeeding: what to consider. Current Psychiatry. 2023;22(5):20-23,48. doi:10.12788/cp.0355

Drug Brand Names

Aripiprazole • Abilify
Olanzapine • Zyprexa
Risperidone • Risperdal

In the United States, opioid use by patients who are pregnant more than quadrupled from 1999 to 2014.¹ Opioid use disorder (OUD) in the perinatal period is associated with a higher risk for depression, suicide, malnutrition, domestic violence, and obstetric complications such as spontaneous abortion, preeclampsia, and premature delivery.² Buprenorphine and methadone are the standard of care for treating OUD in pregnancy.^3,4 While a literature review found that maternal treatment with buprenorphine has comparable efficacy to treatment with methadone,⁵ a small randomized, double-blind study found that compared to buprenorphine, methadone was associated with significantly lower use of additional opioids (P = .047).⁶ This suggests methadone has therapeutic value for patients who are pregnant.

Despite the benefits of methadone for treating perinatal OUD, the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Patients typically take methadone once a day, and the dose is titrated every 3 to 5 days to allow serum levels to reach steady state.⁷ During pregnancy, there are increases in both the volume of distribution and medication metabolism secondary to increased expression of the cytochrome P450 3A4 enzyme by the liver, intestine, and placenta.⁸ Additionally, as the pregnancy progresses, the rate of methadone metabolism increases.⁹ Methadone’s half-life (20 to 35 hours) leads to its accumulation in tissue and slow release into the blood.¹⁰ As a result, patients with OUD who are pregnant often require higher doses of methadone or divided dosing, particularly in the second and third trimesters.¹¹

In this article, we provide a strategy for divided dosing of methadone for managing opioid withdrawal symptoms in the acute care setting. We present 2 cases of women with OUD who are pregnant and describe the collaboration of addiction medicine, consultation-liaison psychiatry, and obstetrics services.

CASE 1

Ms. H, age 29, is G3P2 and presents to the emergency department (ED) during her fourth pregnancy at 31 weeks, 1 day gestation. She has a history of opioid, cocaine, and benzodiazepine use disorders and chronic hepatitis C. Ms. H is enrolled in an opioid treatment program and takes methadone 190 mg/d in addition to nonprescribed opioids. In the ED, Ms. H requests medically supervised withdrawal management. Her urine toxicology is positive for cocaine, benzodiazepines, methadone, and opiates. Her laboratory results and electrocardiogram (ECG) are unremarkable. On admission, Ms. H’s Clinical Opiate Withdrawal Scale (COWS) score is 3, indicating minimal symptoms (5 to 12: mild; 13 to 24: moderate; 25 to 36: moderately severe; >36: severe). Fetal monitoring is reassuring.

Ms. H’s withdrawal is monitored with COWS every 4 hours. The treatment team initiates methadone 170 mg/d, with an additional 10 mg/d as needed to keep her COWS score <8, and daily QTc monitoring. Ms. H also receives lorazepam 2 to 4 mg/d as needed for benzodiazepine withdrawal. Despite the increase in her daily methadone dose, Ms. H continues to experience opioid withdrawal in the early evening and overnight. As a result, the treatment team increases Ms. H’s morning methadone dose to 190 mg and schedules an afternoon dose of 30 mg. Despite this adjustment, her COWS scores remain elevated in the afternoon and evening, and she requires additional as-needed doses of methadone. Methadone peak and trough levels are ordered to assess for rapid metabolism. The serum trough level is 190 ng/mL, which is low, and a serum peak level is not reported. Despite titration, Ms. H has a self-directed premature discharge.

Five days later at 32 weeks, 2 days gestation, Ms. H is readmitted after she had resumed use of opioids, benzodiazepines, and cocaine. Her vital signs are stable, and her laboratory results and ECG are unremarkable. Fetal monitoring is reassuring. Given Ms. H’s low methadone serum trough level and overall concern for rapid methadone metabolism, the treatment team decides to divide dosing of methadone. Over 9 days, the team titrates methadone to 170 mg twice daily on the day of discharge, which resolves Ms. H’s withdrawal symptoms.

At 38 weeks, 5 days gestation, Ms. H returns to the ED after experiencing labor contractions and opiate withdrawal symptoms after she resumed use of heroin, cocaine, and benzodiazepines. During this admission, Ms. H’s methadone is increased to 180 mg twice daily with additional as-needed doses for ongoing withdrawal symptoms. At 39 weeks, 2 days gestation, Ms. H has a scheduled cesarean delivery.

Her infant has a normal weight but is transferred to the neonatal intensive care unit (NICU) for management of neonatal opioid withdrawal syndrome (NOWS) and receives morphine. The baby remains in the NICU for 35 days and is discharged home without further treatment. When Ms. H is discharged, her methadone dose is 170 mg twice daily, which resolves her opioid withdrawal symptoms. The treatment team directs her to continue care in her methadone outpatient program and receive treatment for her cocaine and benzodiazepine use disorders. She declines residential or inpatient substance use treatment.

Continue to: CASE 2

Ms. M, age 39, is G4P2 and presents to the hospital during her fifth pregnancy at 27 weeks gestation. She has not received prenatal care for this pregnancy. She has a history of OUD and major depressive disorder (MDD). Ms. M’s urine toxicology is positive for opiates, fentanyl, and oxycodone. Her laboratory results are notable for mildly elevated alanine aminotransferase, positive hepatitis C antibody, and a hepatitis C viral load of 91,000, consistent with chronic hepatitis C infection. On admission, her COWS score is 14, indicating moderate withdrawal symptoms. Her ECG is unremarkable, and fetal monitoring is reassuring.

Ms. M had received methadone during a prior pregnancy and opts to reinitiate treatment with methadone during her current admission. The team initiates methadone 20 mg/d with additional as-needed doses for ongoing withdrawal symptoms. Due to a persistently elevated COWS score, Ms. M’s methadone is increased to 90 mg/d, which resolves her withdrawal symptoms. However, on Day 4, Ms. M reports having anxiety, refuses bloodwork to obtain methadone peak and trough levels, and prematurely discharges from the hospital.

One day later at 27 weeks, 5 days gestation, Ms. M is readmitted for continued management of opioid withdrawal. She presents with stable vital signs, an unremarkable ECG, and reassuring fetal monitoring. Her COWS score is 5. The treatment team reinitiates methadone at 80 mg/d and titrates it to 100 mg/d on Day 7. Given Ms. M’s ongoing evening cravings and concern for rapid methadone metabolism, on Day 10 the team switches the methadone dosing to 50 mg twice daily to maintain steady-state levels and promote patient comfort. Fluoxetine 20 mg/d is started for comorbid MDD and eventually increased to 80 mg/d. Ms. M is discharged on Day 15 with a regimen of methadone 60 mg/d in the morning and 70 mg/d at night. She plans to resume care in an opioid treatment program and follow up with psychiatry and hepatology for her anxiety and hepatitis C.

A need for aggressive treatment

Given the rising rates of opioid use by patients who are pregnant, harmful behavior related to opioid use, and a wealth of evidence supporting opioid agonist treatment for OUD in pregnancy, there is a growing need for guidance in managing perinatal OUD. A systematic approach to using methadone to treat OUD in patients who are pregnant is essential; the lack of data surrounding use of this medication in such patients may cause overall harm.¹² Limited guidelines and a lack of familiarity with prescribing methadone to patients who are pregnant may lead clinicians to underdose patients, which can result in ongoing withdrawal, premature patient-directed discharges, and poor engagement in care.¹³ Both patients in the 2 cases described in this article experienced ongoing withdrawal symptoms despite daily titration of methadone. This suggests rapid metabolism, which was successfully managed by dividing the dosing of methadone, particularly in the latter trimesters.

These cases illustrate the need for aggressive perinatal opioid withdrawal management through rapid escalation of divided doses of methadone in a monitored acute care setting. Because methadone elimination is more rapid and clearance rates increase during the perinatal period, divided methadone dosing allows for sustained plasma methadone concentrations and improved outpatient treatment adherence.^9,14,15

Continue to: Decreasing the rate of premature discharges

Decreasing the rate of premature discharges

In both cases, the patients discharged from the hospital prematurely, likely related to incomplete management of their opioid withdrawal or other withdrawal syndromes (both patients had multiple substance use disorders [SUDs]). Compared to patients without an SUD, patients with SUDs are 3 times more likely to have a self-directed discharge.¹⁶ Patients report leaving the hospital prematurely due to undertreated withdrawal, uncontrolled pain, discrimination by staff, and hospital restrictions.¹⁶ Recommendations to decrease the rates of premature patient-directed discharges in this population include providing patient-centered and harm reduction–oriented care in addition to adequate management of pain and withdrawal.¹⁷

Impact of methadone on fetal outcomes

Approximately 55% to 94% of infants born to patients who are opioid-dependent will develop NOWS. However, there is no relationship between this syndrome and therapeutic doses of methadone.¹⁸ Moreover, long-term research has found that after adjusting for socioeconomic factors, methadone treatment during pregnancy does not have an adverse effect on postnatal development. Divided dosing in maternal methadone administration is also shown to have less of an impact on fetal neurobehavior and NOWS.¹⁹

Our recommendations for methadone treatment for perinatal patients are outlined in the Table. Aggressive treatment of opioid withdrawal in the hospital can promote treatment engagement and prevent premature discharges. Clinicians should assess for other withdrawal syndromes when a patient has multiple SUDs and collaborate with an interdisciplinary team to improve patient outcomes.

Bottom Line

The prevalence of opioid use disorder (OUD) in patients who are pregnant is increasing. Methadone is an option for treating perinatal OUD, but the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Using divided doses of methadone can ensure the comfort and safety of the patient and their baby and improve adherence and outcomes.

Related Resources

• Chaney L, Mathia C, Cole T. Transitioning patients with opioid use disorder from methadone to buprenorphine. Current Psychiatry. 2022;21(12):23-24,28. doi:10.12788/cp.0305
• Townsel C, Irani S, Buis C, et al. Partnering for the future clinic: a multidisciplinary perinatal substance use program. Gen Hosp Psychiatry. 2023;85:220-228. doi:10.1016/j. genhosppsych.2023.10.009

Drug Brand Names

Buprenorphine • Buprenex, Suboxone, Zubsolv, Sublocade
Fentanyl • Abstral, Actiq
Fluoxetine • Prozac
Lorazepam • Ativan
Methadone • Methadose, Dolophine
Oxycodone • Oxycontin

In the United States, opioid use by patients who are pregnant more than quadrupled from 1999 to 2014.¹ Opioid use disorder (OUD) in the perinatal period is associated with a higher risk for depression, suicide, malnutrition, domestic violence, and obstetric complications such as spontaneous abortion, preeclampsia, and premature delivery.² Buprenorphine and methadone are the standard of care for treating OUD in pregnancy.^3,4 While a literature review found that maternal treatment with buprenorphine has comparable efficacy to treatment with methadone,⁵ a small randomized, double-blind study found that compared to buprenorphine, methadone was associated with significantly lower use of additional opioids (P = .047).⁶ This suggests methadone has therapeutic value for patients who are pregnant.

Despite the benefits of methadone for treating perinatal OUD, the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Patients typically take methadone once a day, and the dose is titrated every 3 to 5 days to allow serum levels to reach steady state.⁷ During pregnancy, there are increases in both the volume of distribution and medication metabolism secondary to increased expression of the cytochrome P450 3A4 enzyme by the liver, intestine, and placenta.⁸ Additionally, as the pregnancy progresses, the rate of methadone metabolism increases.⁹ Methadone’s half-life (20 to 35 hours) leads to its accumulation in tissue and slow release into the blood.¹⁰ As a result, patients with OUD who are pregnant often require higher doses of methadone or divided dosing, particularly in the second and third trimesters.¹¹

In this article, we provide a strategy for divided dosing of methadone for managing opioid withdrawal symptoms in the acute care setting. We present 2 cases of women with OUD who are pregnant and describe the collaboration of addiction medicine, consultation-liaison psychiatry, and obstetrics services.

CASE 1

Ms. H, age 29, is G3P2 and presents to the emergency department (ED) during her fourth pregnancy at 31 weeks, 1 day gestation. She has a history of opioid, cocaine, and benzodiazepine use disorders and chronic hepatitis C. Ms. H is enrolled in an opioid treatment program and takes methadone 190 mg/d in addition to nonprescribed opioids. In the ED, Ms. H requests medically supervised withdrawal management. Her urine toxicology is positive for cocaine, benzodiazepines, methadone, and opiates. Her laboratory results and electrocardiogram (ECG) are unremarkable. On admission, Ms. H’s Clinical Opiate Withdrawal Scale (COWS) score is 3, indicating minimal symptoms (5 to 12: mild; 13 to 24: moderate; 25 to 36: moderately severe; >36: severe). Fetal monitoring is reassuring.

Ms. H’s withdrawal is monitored with COWS every 4 hours. The treatment team initiates methadone 170 mg/d, with an additional 10 mg/d as needed to keep her COWS score <8, and daily QTc monitoring. Ms. H also receives lorazepam 2 to 4 mg/d as needed for benzodiazepine withdrawal. Despite the increase in her daily methadone dose, Ms. H continues to experience opioid withdrawal in the early evening and overnight. As a result, the treatment team increases Ms. H’s morning methadone dose to 190 mg and schedules an afternoon dose of 30 mg. Despite this adjustment, her COWS scores remain elevated in the afternoon and evening, and she requires additional as-needed doses of methadone. Methadone peak and trough levels are ordered to assess for rapid metabolism. The serum trough level is 190 ng/mL, which is low, and a serum peak level is not reported. Despite titration, Ms. H has a self-directed premature discharge.

Five days later at 32 weeks, 2 days gestation, Ms. H is readmitted after she had resumed use of opioids, benzodiazepines, and cocaine. Her vital signs are stable, and her laboratory results and ECG are unremarkable. Fetal monitoring is reassuring. Given Ms. H’s low methadone serum trough level and overall concern for rapid methadone metabolism, the treatment team decides to divide dosing of methadone. Over 9 days, the team titrates methadone to 170 mg twice daily on the day of discharge, which resolves Ms. H’s withdrawal symptoms.

At 38 weeks, 5 days gestation, Ms. H returns to the ED after experiencing labor contractions and opiate withdrawal symptoms after she resumed use of heroin, cocaine, and benzodiazepines. During this admission, Ms. H’s methadone is increased to 180 mg twice daily with additional as-needed doses for ongoing withdrawal symptoms. At 39 weeks, 2 days gestation, Ms. H has a scheduled cesarean delivery.

Her infant has a normal weight but is transferred to the neonatal intensive care unit (NICU) for management of neonatal opioid withdrawal syndrome (NOWS) and receives morphine. The baby remains in the NICU for 35 days and is discharged home without further treatment. When Ms. H is discharged, her methadone dose is 170 mg twice daily, which resolves her opioid withdrawal symptoms. The treatment team directs her to continue care in her methadone outpatient program and receive treatment for her cocaine and benzodiazepine use disorders. She declines residential or inpatient substance use treatment.

Continue to: CASE 2

Ms. M, age 39, is G4P2 and presents to the hospital during her fifth pregnancy at 27 weeks gestation. She has not received prenatal care for this pregnancy. She has a history of OUD and major depressive disorder (MDD). Ms. M’s urine toxicology is positive for opiates, fentanyl, and oxycodone. Her laboratory results are notable for mildly elevated alanine aminotransferase, positive hepatitis C antibody, and a hepatitis C viral load of 91,000, consistent with chronic hepatitis C infection. On admission, her COWS score is 14, indicating moderate withdrawal symptoms. Her ECG is unremarkable, and fetal monitoring is reassuring.

Ms. M had received methadone during a prior pregnancy and opts to reinitiate treatment with methadone during her current admission. The team initiates methadone 20 mg/d with additional as-needed doses for ongoing withdrawal symptoms. Due to a persistently elevated COWS score, Ms. M’s methadone is increased to 90 mg/d, which resolves her withdrawal symptoms. However, on Day 4, Ms. M reports having anxiety, refuses bloodwork to obtain methadone peak and trough levels, and prematurely discharges from the hospital.

One day later at 27 weeks, 5 days gestation, Ms. M is readmitted for continued management of opioid withdrawal. She presents with stable vital signs, an unremarkable ECG, and reassuring fetal monitoring. Her COWS score is 5. The treatment team reinitiates methadone at 80 mg/d and titrates it to 100 mg/d on Day 7. Given Ms. M’s ongoing evening cravings and concern for rapid methadone metabolism, on Day 10 the team switches the methadone dosing to 50 mg twice daily to maintain steady-state levels and promote patient comfort. Fluoxetine 20 mg/d is started for comorbid MDD and eventually increased to 80 mg/d. Ms. M is discharged on Day 15 with a regimen of methadone 60 mg/d in the morning and 70 mg/d at night. She plans to resume care in an opioid treatment program and follow up with psychiatry and hepatology for her anxiety and hepatitis C.

A need for aggressive treatment

Given the rising rates of opioid use by patients who are pregnant, harmful behavior related to opioid use, and a wealth of evidence supporting opioid agonist treatment for OUD in pregnancy, there is a growing need for guidance in managing perinatal OUD. A systematic approach to using methadone to treat OUD in patients who are pregnant is essential; the lack of data surrounding use of this medication in such patients may cause overall harm.¹² Limited guidelines and a lack of familiarity with prescribing methadone to patients who are pregnant may lead clinicians to underdose patients, which can result in ongoing withdrawal, premature patient-directed discharges, and poor engagement in care.¹³ Both patients in the 2 cases described in this article experienced ongoing withdrawal symptoms despite daily titration of methadone. This suggests rapid metabolism, which was successfully managed by dividing the dosing of methadone, particularly in the latter trimesters.

These cases illustrate the need for aggressive perinatal opioid withdrawal management through rapid escalation of divided doses of methadone in a monitored acute care setting. Because methadone elimination is more rapid and clearance rates increase during the perinatal period, divided methadone dosing allows for sustained plasma methadone concentrations and improved outpatient treatment adherence.^9,14,15

Continue to: Decreasing the rate of premature discharges

Decreasing the rate of premature discharges

In both cases, the patients discharged from the hospital prematurely, likely related to incomplete management of their opioid withdrawal or other withdrawal syndromes (both patients had multiple substance use disorders [SUDs]). Compared to patients without an SUD, patients with SUDs are 3 times more likely to have a self-directed discharge.¹⁶ Patients report leaving the hospital prematurely due to undertreated withdrawal, uncontrolled pain, discrimination by staff, and hospital restrictions.¹⁶ Recommendations to decrease the rates of premature patient-directed discharges in this population include providing patient-centered and harm reduction–oriented care in addition to adequate management of pain and withdrawal.¹⁷

Impact of methadone on fetal outcomes

Approximately 55% to 94% of infants born to patients who are opioid-dependent will develop NOWS. However, there is no relationship between this syndrome and therapeutic doses of methadone.¹⁸ Moreover, long-term research has found that after adjusting for socioeconomic factors, methadone treatment during pregnancy does not have an adverse effect on postnatal development. Divided dosing in maternal methadone administration is also shown to have less of an impact on fetal neurobehavior and NOWS.¹⁹

Our recommendations for methadone treatment for perinatal patients are outlined in the Table. Aggressive treatment of opioid withdrawal in the hospital can promote treatment engagement and prevent premature discharges. Clinicians should assess for other withdrawal syndromes when a patient has multiple SUDs and collaborate with an interdisciplinary team to improve patient outcomes.

Bottom Line

The prevalence of opioid use disorder (OUD) in patients who are pregnant is increasing. Methadone is an option for treating perinatal OUD, but the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Using divided doses of methadone can ensure the comfort and safety of the patient and their baby and improve adherence and outcomes.

Related Resources

• Chaney L, Mathia C, Cole T. Transitioning patients with opioid use disorder from methadone to buprenorphine. Current Psychiatry. 2022;21(12):23-24,28. doi:10.12788/cp.0305
• Townsel C, Irani S, Buis C, et al. Partnering for the future clinic: a multidisciplinary perinatal substance use program. Gen Hosp Psychiatry. 2023;85:220-228. doi:10.1016/j. genhosppsych.2023.10.009

Drug Brand Names

Buprenorphine • Buprenex, Suboxone, Zubsolv, Sublocade
Fentanyl • Abstral, Actiq
Fluoxetine • Prozac
Lorazepam • Ativan
Methadone • Methadose, Dolophine
Oxycodone • Oxycontin

In the United States, opioid use by patients who are pregnant more than quadrupled from 1999 to 2014.¹ Opioid use disorder (OUD) in the perinatal period is associated with a higher risk for depression, suicide, malnutrition, domestic violence, and obstetric complications such as spontaneous abortion, preeclampsia, and premature delivery.² Buprenorphine and methadone are the standard of care for treating OUD in pregnancy.^3,4 While a literature review found that maternal treatment with buprenorphine has comparable efficacy to treatment with methadone,⁵ a small randomized, double-blind study found that compared to buprenorphine, methadone was associated with significantly lower use of additional opioids (P = .047).⁶ This suggests methadone has therapeutic value for patients who are pregnant.

Despite the benefits of methadone for treating perinatal OUD, the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Patients typically take methadone once a day, and the dose is titrated every 3 to 5 days to allow serum levels to reach steady state.⁷ During pregnancy, there are increases in both the volume of distribution and medication metabolism secondary to increased expression of the cytochrome P450 3A4 enzyme by the liver, intestine, and placenta.⁸ Additionally, as the pregnancy progresses, the rate of methadone metabolism increases.⁹ Methadone’s half-life (20 to 35 hours) leads to its accumulation in tissue and slow release into the blood.¹⁰ As a result, patients with OUD who are pregnant often require higher doses of methadone or divided dosing, particularly in the second and third trimesters.¹¹

In this article, we provide a strategy for divided dosing of methadone for managing opioid withdrawal symptoms in the acute care setting. We present 2 cases of women with OUD who are pregnant and describe the collaboration of addiction medicine, consultation-liaison psychiatry, and obstetrics services.

CASE 1

Ms. H, age 29, is G3P2 and presents to the emergency department (ED) during her fourth pregnancy at 31 weeks, 1 day gestation. She has a history of opioid, cocaine, and benzodiazepine use disorders and chronic hepatitis C. Ms. H is enrolled in an opioid treatment program and takes methadone 190 mg/d in addition to nonprescribed opioids. In the ED, Ms. H requests medically supervised withdrawal management. Her urine toxicology is positive for cocaine, benzodiazepines, methadone, and opiates. Her laboratory results and electrocardiogram (ECG) are unremarkable. On admission, Ms. H’s Clinical Opiate Withdrawal Scale (COWS) score is 3, indicating minimal symptoms (5 to 12: mild; 13 to 24: moderate; 25 to 36: moderately severe; >36: severe). Fetal monitoring is reassuring.

Ms. H’s withdrawal is monitored with COWS every 4 hours. The treatment team initiates methadone 170 mg/d, with an additional 10 mg/d as needed to keep her COWS score <8, and daily QTc monitoring. Ms. H also receives lorazepam 2 to 4 mg/d as needed for benzodiazepine withdrawal. Despite the increase in her daily methadone dose, Ms. H continues to experience opioid withdrawal in the early evening and overnight. As a result, the treatment team increases Ms. H’s morning methadone dose to 190 mg and schedules an afternoon dose of 30 mg. Despite this adjustment, her COWS scores remain elevated in the afternoon and evening, and she requires additional as-needed doses of methadone. Methadone peak and trough levels are ordered to assess for rapid metabolism. The serum trough level is 190 ng/mL, which is low, and a serum peak level is not reported. Despite titration, Ms. H has a self-directed premature discharge.

Five days later at 32 weeks, 2 days gestation, Ms. H is readmitted after she had resumed use of opioids, benzodiazepines, and cocaine. Her vital signs are stable, and her laboratory results and ECG are unremarkable. Fetal monitoring is reassuring. Given Ms. H’s low methadone serum trough level and overall concern for rapid methadone metabolism, the treatment team decides to divide dosing of methadone. Over 9 days, the team titrates methadone to 170 mg twice daily on the day of discharge, which resolves Ms. H’s withdrawal symptoms.

At 38 weeks, 5 days gestation, Ms. H returns to the ED after experiencing labor contractions and opiate withdrawal symptoms after she resumed use of heroin, cocaine, and benzodiazepines. During this admission, Ms. H’s methadone is increased to 180 mg twice daily with additional as-needed doses for ongoing withdrawal symptoms. At 39 weeks, 2 days gestation, Ms. H has a scheduled cesarean delivery.

Her infant has a normal weight but is transferred to the neonatal intensive care unit (NICU) for management of neonatal opioid withdrawal syndrome (NOWS) and receives morphine. The baby remains in the NICU for 35 days and is discharged home without further treatment. When Ms. H is discharged, her methadone dose is 170 mg twice daily, which resolves her opioid withdrawal symptoms. The treatment team directs her to continue care in her methadone outpatient program and receive treatment for her cocaine and benzodiazepine use disorders. She declines residential or inpatient substance use treatment.

Continue to: CASE 2

Ms. M, age 39, is G4P2 and presents to the hospital during her fifth pregnancy at 27 weeks gestation. She has not received prenatal care for this pregnancy. She has a history of OUD and major depressive disorder (MDD). Ms. M’s urine toxicology is positive for opiates, fentanyl, and oxycodone. Her laboratory results are notable for mildly elevated alanine aminotransferase, positive hepatitis C antibody, and a hepatitis C viral load of 91,000, consistent with chronic hepatitis C infection. On admission, her COWS score is 14, indicating moderate withdrawal symptoms. Her ECG is unremarkable, and fetal monitoring is reassuring.

Ms. M had received methadone during a prior pregnancy and opts to reinitiate treatment with methadone during her current admission. The team initiates methadone 20 mg/d with additional as-needed doses for ongoing withdrawal symptoms. Due to a persistently elevated COWS score, Ms. M’s methadone is increased to 90 mg/d, which resolves her withdrawal symptoms. However, on Day 4, Ms. M reports having anxiety, refuses bloodwork to obtain methadone peak and trough levels, and prematurely discharges from the hospital.

One day later at 27 weeks, 5 days gestation, Ms. M is readmitted for continued management of opioid withdrawal. She presents with stable vital signs, an unremarkable ECG, and reassuring fetal monitoring. Her COWS score is 5. The treatment team reinitiates methadone at 80 mg/d and titrates it to 100 mg/d on Day 7. Given Ms. M’s ongoing evening cravings and concern for rapid methadone metabolism, on Day 10 the team switches the methadone dosing to 50 mg twice daily to maintain steady-state levels and promote patient comfort. Fluoxetine 20 mg/d is started for comorbid MDD and eventually increased to 80 mg/d. Ms. M is discharged on Day 15 with a regimen of methadone 60 mg/d in the morning and 70 mg/d at night. She plans to resume care in an opioid treatment program and follow up with psychiatry and hepatology for her anxiety and hepatitis C.

A need for aggressive treatment

Given the rising rates of opioid use by patients who are pregnant, harmful behavior related to opioid use, and a wealth of evidence supporting opioid agonist treatment for OUD in pregnancy, there is a growing need for guidance in managing perinatal OUD. A systematic approach to using methadone to treat OUD in patients who are pregnant is essential; the lack of data surrounding use of this medication in such patients may cause overall harm.¹² Limited guidelines and a lack of familiarity with prescribing methadone to patients who are pregnant may lead clinicians to underdose patients, which can result in ongoing withdrawal, premature patient-directed discharges, and poor engagement in care.¹³ Both patients in the 2 cases described in this article experienced ongoing withdrawal symptoms despite daily titration of methadone. This suggests rapid metabolism, which was successfully managed by dividing the dosing of methadone, particularly in the latter trimesters.

These cases illustrate the need for aggressive perinatal opioid withdrawal management through rapid escalation of divided doses of methadone in a monitored acute care setting. Because methadone elimination is more rapid and clearance rates increase during the perinatal period, divided methadone dosing allows for sustained plasma methadone concentrations and improved outpatient treatment adherence.^9,14,15

Continue to: Decreasing the rate of premature discharges

Decreasing the rate of premature discharges

In both cases, the patients discharged from the hospital prematurely, likely related to incomplete management of their opioid withdrawal or other withdrawal syndromes (both patients had multiple substance use disorders [SUDs]). Compared to patients without an SUD, patients with SUDs are 3 times more likely to have a self-directed discharge.¹⁶ Patients report leaving the hospital prematurely due to undertreated withdrawal, uncontrolled pain, discrimination by staff, and hospital restrictions.¹⁶ Recommendations to decrease the rates of premature patient-directed discharges in this population include providing patient-centered and harm reduction–oriented care in addition to adequate management of pain and withdrawal.¹⁷

Impact of methadone on fetal outcomes

Approximately 55% to 94% of infants born to patients who are opioid-dependent will develop NOWS. However, there is no relationship between this syndrome and therapeutic doses of methadone.¹⁸ Moreover, long-term research has found that after adjusting for socioeconomic factors, methadone treatment during pregnancy does not have an adverse effect on postnatal development. Divided dosing in maternal methadone administration is also shown to have less of an impact on fetal neurobehavior and NOWS.¹⁹

Our recommendations for methadone treatment for perinatal patients are outlined in the Table. Aggressive treatment of opioid withdrawal in the hospital can promote treatment engagement and prevent premature discharges. Clinicians should assess for other withdrawal syndromes when a patient has multiple SUDs and collaborate with an interdisciplinary team to improve patient outcomes.

Bottom Line

The prevalence of opioid use disorder (OUD) in patients who are pregnant is increasing. Methadone is an option for treating perinatal OUD, but the physiological changes that occur in patients who are pregnant—coupled with methadone’s unique pharmacologic properties—may complicate its use. Using divided doses of methadone can ensure the comfort and safety of the patient and their baby and improve adherence and outcomes.

Related Resources

• Chaney L, Mathia C, Cole T. Transitioning patients with opioid use disorder from methadone to buprenorphine. Current Psychiatry. 2022;21(12):23-24,28. doi:10.12788/cp.0305
• Townsel C, Irani S, Buis C, et al. Partnering for the future clinic: a multidisciplinary perinatal substance use program. Gen Hosp Psychiatry. 2023;85:220-228. doi:10.1016/j. genhosppsych.2023.10.009

Drug Brand Names

Buprenorphine • Buprenex, Suboxone, Zubsolv, Sublocade
Fentanyl • Abstral, Actiq
Fluoxetine • Prozac
Lorazepam • Ativan
Methadone • Methadose, Dolophine
Oxycodone • Oxycontin

Delirious mania is a syndrome characterized by the acute onset of severe hyperactivity, psychosis, catatonia, and intermittent confusion. While there have been growing reports of this phenomenon over the last 2 decades, it remains poorly recognized and understood.^1,2 There is no widely accepted nosology for delirious mania and the condition is absent from DSM-5, which magnifies the difficulties in making a timely diagnosis and initiating appropriate treatment. Delayed diagnosis and treatment may result in a detrimental outcome.^2,3 Delirious mania has also been labeled as lethal catatonia, specific febrile delirium, hyperactive or exhaustive mania, and Bell’s mania.^2,4,5 The characterization and diagnosis of this condition have a long and inconsistent history (Box^1,6-11).

Box

Delirious mania continues to be met with incertitude in clinical practice, and numerous inconsistencies have been reported in the literature. For example, some cases that have been reported as delirious mania had more evidence of primary delirium due to another medical condition or primary mania.^12,13 Other cases have demonstrated swift improvement of symptoms after monotherapy with antipsychotics without a trial of benzodiazepines or electroconvulsive therapy (ECT); the exclusion of a sudden onset questions the validity of the diagnosis and promotes the use of less efficacious treatments.^14,15 Other reports have confirmed that the diagnosis is missed when certain symptoms are more predominant, such as a thought disorder (acute schizophrenia), grandiosity and delusional ideation (bipolar disorder [BD]), and less commonly assessed catatonic signs (ambitendency, automatic obedience). These symptoms are mistakenly attributed to the respective disease.^1,16 This especially holds true when delirious mania is initially diagnosed as a primary psychosis, which leads to the administration of antipsychotics.¹⁷ Other cases have reported that delirious mania was resistant to treatment, but ECT was never pursued.¹⁸

In this review, we provide a more comprehensive perspective of the clinical presentation, pathogenesis, and management of delirious mania. We searched PubMed and Google Scholar using the keywords “delirious mania,” “delirious mania AND catatonia,” or “manic delirium.” Most articles we found were case reports, case series, or retrospective chart reviews. There were no systematic reviews, meta analyses, or randomized control trials (RCTs). The 12 articles included in this review consist of 7 individual case reports, 4 case series, and 1 retrospective chart review that describe a total of 36 cases (Table^{1,2,5,17,19-26}).

Clinical presentation: What to look for

Patients with delirious mania typically develop symptoms extremely rapidly. In virtually all published literature, symptoms were reported to emerge within hours to days and consisted of severe forms of mania, psychosis, and delirium; 100% of the cases in our review had these symptoms. Commonly reported symptoms were:

• intense excitement
• emotional lability
• grandiose delusions
• profound insomnia
• pressured and rapid speech
• auditory and visual hallucinations
• hypersexuality
• thought disorganization.

Exquisite paranoia can also result in violent aggression (and may require the use of physical restraints). Patients may confine themselves to very small spaces (such as a closet) in response to the intense paranoia. Impairments in various neurocognitive domains—including inability to focus; disorientation; language and visuospatial disturbances; difficulty with shifting and sustaining attention; and short-term memory impairments—have been reported. Patients often cannot recall the events during the episode.^1,2,5,27,28

Catatonia has been closely associated with delirious mania.²⁹ Features of excited catatonia—such as excessive motor activity, negativism, grimacing, posturing, echolalia, echopraxia, stereotypy, automatic obedience, verbigeration, combativeness, impulsivity, and rigidity—typically accompany delirious mania.^1,5,10,19,27

In addition to these symptoms, patients may engage in specific behaviors. They may exhibit inappropriate toileting such as smearing feces on walls or in bags, fecal or urinary incontinence, disrobing or running naked in public places, or pouring liquid on the floor or on one’s head.^1,2

Continue to: Of the 36 cases...

Of the 36 cases reported in the literature we reviewed, 20 (55%) were female. Most patients had an underlining psychiatric condition, including BD (72%), major depressive disorder (8%), and schizophrenia (2%). Three patients had no psychiatric history.

Physical examination

On initial presentation, a patient with delirious mania may be dehydrated, with dry mucous membranes, pale conjunctiva, tongue dryness, and poor skin turgor.^28,30 Due to excessive motor activity, diaphoresis with tachycardia, fluctuating blood pressure, and fever may be present.³¹

Certain basic cognitive tasks should be assessed to determine the patient’s orientation to place, date, and time. Assess if the patient can recall recent events, names of objects, or perform serial 7s; clock drawing capabilities also should be ascertained.^1,2,5 A Mini-Mental State Examination is useful.³²

The Bush-Francis Catatonia Rating Scale should be used to elicit features of catatonia, such as waxy flexibility, negativism, gegenhalten, mitgehen, catalepsy, ambitendency, automatic obedience, and grasp reflex.¹⁰

Laboratory findings are nonspecific

Although no specific laboratory findings are associated with delirious mania, bloodwork and imaging are routinely investigated, especially if delirium characteristics are most striking. A complete blood count, chemistries, hepatic panel, thyroid functioning, blood and/or urine cultures, creatinine phosphokinase (CPK), and urinalysis can be ordered. Head imaging such as MRI and CT to rule out intracranial pathology are typically performed.¹⁹ However, the diagnosis of delirious mania is based on the presence of the phenotypic features, by verification of catatonia, and by the responsiveness to the treatment delivered.²⁹

Continue to: Pathogenisis: Several hypotheses

The pathogenesis of delirious mania is not well understood. There are several postulations but no salient theory. Most patients with delirious mania have an underlying systemic medical or psychiatric condition.

Mood disorders. Patients with BD or schizoaffective disorder are especially susceptible to delirious mania. The percentage of manic patients who present with delirious mania varies by study. One study suggested approximately 19% have features of the phenomenon,³³ while others estimated 15% to 25%.³⁴ Elias et al³⁵ calculated that 15% of patients with mania succumb to manic exhaustion; from this it can be reasonably concluded that these were cases of misdiagnosed delirious mania.

Delirium hypothesis. Patients with delirious mania typically have features of delirium, including fluctuation of consciousness, disorientation, and/or poor sleep-wake cycle.³⁶ During rapid eye movement (REM) and non-REM sleep, memory circuits are fortified. When there is a substantial loss of REM and non-REM sleep, these circuits become faulty, even after 1 night. Pathological brain waves on EEG reflect the inability to reinforce the memory circuits. Patients with these waves may develop hallucinations, bizarre delusions, and altered sensorium. ECT reduces the pathological slow wave morphologies, thus restoring the synaptic maintenance and correcting the incompetent circuitry. This can explain the robust and rapid response of ECT in a patient with delirious mania.^37,38

Neurotransmitter hypothesis. It has been shown that in patients with delirious mania there is dysregulation of dopamine transport, which leads to dopamine overflow in the synapse. In contrast to a drug effect (ie, cocaine or methamphetamine) that acts by inhibiting dopamine reuptake, dopamine overflow in delirious mania is caused by the loss of dopamine transporter regulation. This results in a dysfunctional dopaminergic state that precipitates an acute state of delirium and agitation.^39,40

Serotonin plays a role in mood disorders, including mania and depression.^41,42 More specifically, serotonin has been implicated in impulsivity and aggression as shown by reduced levels of CSF 5-hydroxyindoleacetic acid (5-HIAA) and depletion of 5-hydroxytryptophan (5-HTP).⁴³

Continue to: Alterations in gamma-aminobutyric acid (GABA) transmission...

Alterations in gamma-aminobutyric acid (GABA) transmission are known to occur in delirium and catatonia. In delirium, GABA signaling is increased, which disrupts the circadian rhythm and melatonin release, thus impairing the sleep-wake cycle.⁴⁴ Deficiencies in acetylcholine and melatonin are seen as well as excess of other neurotransmitters, including norepinephrine and glutamate.⁴⁵ Conversely, in catatonia, functional imaging studies found decreased GABA-A binding in orbito­frontal, prefrontal, parietal, and motor cortical regions.⁴⁶ A study analyzing 10 catatonic patients found decreased density of GABA-A receptors in the left sensorimotor cortex compared to psychiatric and healthy controls.⁴⁷

Other neurotransmitters, such as glutamate, at the N-methyl-D-aspartate receptors (NMDAR) have been hypothesized to be hyperactive, causing downstream dysregulation of GABA functioning.⁴⁸ However, the exact connection between delirious mania and all these receptors and neurotransmitters remains unknown.

Encephalitis hypothesis. The relationship between delirious mania and autoimmune encephalitis suggests delirious mania has etiologies other than a primary psychiatric illness. In a 2020 retrospective study⁴⁹ that analyzed 79 patients with anti-NMDAR encephalitis, 25.3% met criteria for delirious mania, and 95% of these patients had catatonic features. Dalmau et al⁵⁰ found that in many cases, patients tend to respond to ECT; in a cases series of 3 patients, 2 responded to benzodiazepines.

COVID-19 hypothesis. The SARS-CoV-2 virion has been associated with many neuropsychiatric complications, including mood, psychotic, and neurocognitive disorders.^51,52 There also have been cases of COVID-19–induced catatonia.^53-55 One case of delirious mania in a patient with COVID-19 has been reported.²¹ The general mechanism has been proposed to be related to the stimulation of the proinflammatory cytokines, such as tumor necrosis factor-alpha and interleukin-6, which the virus produces in large quantities.⁵⁶ These cytokines have been linked to psychosis and other psychiatric disorders.⁵⁷ The patient with COVID-19–induced delirious mania had elevated inflammatory markers, including erythrocyte sedimentation rate, C-reactive protein, ferritin, and D-dimer, which supports a proinflammatory state. This patient had a complete resolution of symptoms with ECT.²¹

Management: Benzodiazepines and ECT

A step-by-step algorithm for managing delirious mania is outlined in the Figure. Regardless of the underlining etiology, management of delirious mania consists of benzodiazepines (lorazepam and diazepam); prompt use of ECT, particularly for patients who do not improve with large doses of lorazepam; or (if applicable) continued treatment of the underlining medical condition, which does not preclude the use of benzodiazepines or ECT. Recent reports^27,58 have described details for using ECT for delirious mania, highlighting the use of high-energy dosing, bilateral electrode placement, and frequent sessions.

Continue to: Knowing which medications...

Knowing which medications to avoid is as important as knowing which agents to administer. Be vigilant in avoiding high-potency antipsychotics, as these medications can worsen extrapyramidal symptoms and may precipitate seizures or neuroleptic malignant syndrome (NMS).²⁸ Anticholinergic agents should also be avoided because they worsen confusion. Although lithium is effective in BD, in delirious mania, high doses of lithium and haloperidol may cause severe encephalopathic syndromes, with symptoms that can include lethargy, tremors, cerebellar dysfunction, and worsened confusion; it may also cause widespread and irreversible brain damage.⁵⁹While positive outcomes have been documented when using a combination of antipsychotics and lithium,^8,60 this approach should be considered carefully and tailored to individual cases, taking into account the severity of manic and psychotic symptoms in addition to the level of catatonia.

Due to long periods of hyperactivity, withdrawal, and diaphoresis, patients with delirious mania may be severely dehydrated with metabolic derangements, including elevated CPK due to rhabdomyolysis from prolonged exertion, IM antipsychotics, or rigidity. To prevent acute renal failure, this must be immediately addressed with rapid fluid resuscitation and electrolyte repletion.⁶¹

Benzodiazepines. The rapid use of lorazepam should be initiated when delirious mania is suspected. Doses of 6 to 20 mg have been reported to be effective if tolerated.^5,20 Typically, high-dose lorazepam will not have the sedative effect that would normally occur in a patient who does not have delirious mania.² Lorazepam should be titrated until full resolution of symptoms. Doses up to 30 mg have been reported as effective and tolerable.⁶² In our literature review, 50% of patients (18/36) responded or partially responded to lorazepam. However, only 3 case reports documented a complete remission with lorazepam, and many patients needed ECT for remission of symptoms.

ECT is generally reserved for patients who are not helped by benzodiazepine therapy, which is estimated to be up to 20%.⁵ ECT is highly effective in delirious mania, with remission rates ranging from 80% to 100%.¹ ECT is also effective in acute non­delirious mania (comparable to depression); however, it is only used in a small minority of cases (0.2% to 12%).³⁵ In our review, 58% of cases (21/36) reported using ECT, and in all cases it resulted in complete remission.

A dramatic improvement can be seen even after a single ECT session, though most patients show improvement after 4 sessions or 3 to 7 days.^1,2,5 In our review, most patients received 4 to 12 sessions until achieving complete remission.

Continue to: No RCTs have evaluated...

No RCTs have evaluated ECT electrode placement in patients with delirious mania. However, several RCTs have investigated electrode placement in patients with acute nondelirious mania. Hiremani et al⁶³ found that bitemporal placement had a more rapid response rate than bifrontal placement, but there was no overall difference in response rate. Barekatain et al⁶⁴ found no difference between these 2 bilateral approaches. Many of the delirious mania cases report using a bilateral placement (including 42% of the ECT cases in our review) due to the emergent need for rapid relief of symptoms, which is especially necessary if the patient is experiencing hemodynamic instability, excessive violence, risk for self-harm, worsening delirium, or resistance to lorazepam.

Prognosis: Often fatal if left untreated

Patients with delirious mania are at high risk to progress to a more severe form of NMS or malignant catatonia. Therefore, high-potency antipsychotics should be avoided; mortality can be elevated from 60% without antipsychotics to 78% with antipsychotics.⁴ Some researchers estimate 75% to 78% of cases of delirious mania can be fatal if left untreated.^3,6

Bottom Line

Delirious mania is routinely mistaken for more conventional manic or psychotic disorders. Clinicians need to be able to rapidly recognize the symptoms of this syndrome, which include mania, psychosis, delirium, and possible catatonia, so they can avoid administering toxic agents and instead initiate effective treatments such as benzodiazepines and electroconvulsive therapy.

Related Resources

• Arsan C, Baker C, Wong J, et al. Delirious mania: an approach to diagnosis and treatment. Prim Care Companion CNS Disord. 2021;23(1):20f02744. doi:10.4088/PCC.20f02744
• Lamba G, Kennedy EA, Vu CP. Case report: ECT for delirious mania. Clinical Psychiatry News. December 14, 2021. https://www.mdedge.com/psychiatry/article/249909/bipolar-disorder/case-report-ect-delirious-mania

Drug Brand Names

Diazepam • Valium
Haloperidol • Haldol
Lithium • Eskalith, Lithobid
Lorazepam • Ativan

Delirious mania is a syndrome characterized by the acute onset of severe hyperactivity, psychosis, catatonia, and intermittent confusion. While there have been growing reports of this phenomenon over the last 2 decades, it remains poorly recognized and understood.^1,2 There is no widely accepted nosology for delirious mania and the condition is absent from DSM-5, which magnifies the difficulties in making a timely diagnosis and initiating appropriate treatment. Delayed diagnosis and treatment may result in a detrimental outcome.^2,3 Delirious mania has also been labeled as lethal catatonia, specific febrile delirium, hyperactive or exhaustive mania, and Bell’s mania.^2,4,5 The characterization and diagnosis of this condition have a long and inconsistent history (Box^1,6-11).

Box

Delirious mania continues to be met with incertitude in clinical practice, and numerous inconsistencies have been reported in the literature. For example, some cases that have been reported as delirious mania had more evidence of primary delirium due to another medical condition or primary mania.^12,13 Other cases have demonstrated swift improvement of symptoms after monotherapy with antipsychotics without a trial of benzodiazepines or electroconvulsive therapy (ECT); the exclusion of a sudden onset questions the validity of the diagnosis and promotes the use of less efficacious treatments.^14,15 Other reports have confirmed that the diagnosis is missed when certain symptoms are more predominant, such as a thought disorder (acute schizophrenia), grandiosity and delusional ideation (bipolar disorder [BD]), and less commonly assessed catatonic signs (ambitendency, automatic obedience). These symptoms are mistakenly attributed to the respective disease.^1,16 This especially holds true when delirious mania is initially diagnosed as a primary psychosis, which leads to the administration of antipsychotics.¹⁷ Other cases have reported that delirious mania was resistant to treatment, but ECT was never pursued.¹⁸

In this review, we provide a more comprehensive perspective of the clinical presentation, pathogenesis, and management of delirious mania. We searched PubMed and Google Scholar using the keywords “delirious mania,” “delirious mania AND catatonia,” or “manic delirium.” Most articles we found were case reports, case series, or retrospective chart reviews. There were no systematic reviews, meta analyses, or randomized control trials (RCTs). The 12 articles included in this review consist of 7 individual case reports, 4 case series, and 1 retrospective chart review that describe a total of 36 cases (Table^{1,2,5,17,19-26}).

Clinical presentation: What to look for

Patients with delirious mania typically develop symptoms extremely rapidly. In virtually all published literature, symptoms were reported to emerge within hours to days and consisted of severe forms of mania, psychosis, and delirium; 100% of the cases in our review had these symptoms. Commonly reported symptoms were:

• intense excitement
• emotional lability
• grandiose delusions
• profound insomnia
• pressured and rapid speech
• auditory and visual hallucinations
• hypersexuality
• thought disorganization.

Exquisite paranoia can also result in violent aggression (and may require the use of physical restraints). Patients may confine themselves to very small spaces (such as a closet) in response to the intense paranoia. Impairments in various neurocognitive domains—including inability to focus; disorientation; language and visuospatial disturbances; difficulty with shifting and sustaining attention; and short-term memory impairments—have been reported. Patients often cannot recall the events during the episode.^1,2,5,27,28

Catatonia has been closely associated with delirious mania.²⁹ Features of excited catatonia—such as excessive motor activity, negativism, grimacing, posturing, echolalia, echopraxia, stereotypy, automatic obedience, verbigeration, combativeness, impulsivity, and rigidity—typically accompany delirious mania.^1,5,10,19,27

In addition to these symptoms, patients may engage in specific behaviors. They may exhibit inappropriate toileting such as smearing feces on walls or in bags, fecal or urinary incontinence, disrobing or running naked in public places, or pouring liquid on the floor or on one’s head.^1,2

Continue to: Of the 36 cases...

Of the 36 cases reported in the literature we reviewed, 20 (55%) were female. Most patients had an underlining psychiatric condition, including BD (72%), major depressive disorder (8%), and schizophrenia (2%). Three patients had no psychiatric history.

Physical examination

On initial presentation, a patient with delirious mania may be dehydrated, with dry mucous membranes, pale conjunctiva, tongue dryness, and poor skin turgor.^28,30 Due to excessive motor activity, diaphoresis with tachycardia, fluctuating blood pressure, and fever may be present.³¹

Certain basic cognitive tasks should be assessed to determine the patient’s orientation to place, date, and time. Assess if the patient can recall recent events, names of objects, or perform serial 7s; clock drawing capabilities also should be ascertained.^1,2,5 A Mini-Mental State Examination is useful.³²

The Bush-Francis Catatonia Rating Scale should be used to elicit features of catatonia, such as waxy flexibility, negativism, gegenhalten, mitgehen, catalepsy, ambitendency, automatic obedience, and grasp reflex.¹⁰

Laboratory findings are nonspecific

Although no specific laboratory findings are associated with delirious mania, bloodwork and imaging are routinely investigated, especially if delirium characteristics are most striking. A complete blood count, chemistries, hepatic panel, thyroid functioning, blood and/or urine cultures, creatinine phosphokinase (CPK), and urinalysis can be ordered. Head imaging such as MRI and CT to rule out intracranial pathology are typically performed.¹⁹ However, the diagnosis of delirious mania is based on the presence of the phenotypic features, by verification of catatonia, and by the responsiveness to the treatment delivered.²⁹

Continue to: Pathogenisis: Several hypotheses

The pathogenesis of delirious mania is not well understood. There are several postulations but no salient theory. Most patients with delirious mania have an underlying systemic medical or psychiatric condition.

Mood disorders. Patients with BD or schizoaffective disorder are especially susceptible to delirious mania. The percentage of manic patients who present with delirious mania varies by study. One study suggested approximately 19% have features of the phenomenon,³³ while others estimated 15% to 25%.³⁴ Elias et al³⁵ calculated that 15% of patients with mania succumb to manic exhaustion; from this it can be reasonably concluded that these were cases of misdiagnosed delirious mania.

Delirium hypothesis. Patients with delirious mania typically have features of delirium, including fluctuation of consciousness, disorientation, and/or poor sleep-wake cycle.³⁶ During rapid eye movement (REM) and non-REM sleep, memory circuits are fortified. When there is a substantial loss of REM and non-REM sleep, these circuits become faulty, even after 1 night. Pathological brain waves on EEG reflect the inability to reinforce the memory circuits. Patients with these waves may develop hallucinations, bizarre delusions, and altered sensorium. ECT reduces the pathological slow wave morphologies, thus restoring the synaptic maintenance and correcting the incompetent circuitry. This can explain the robust and rapid response of ECT in a patient with delirious mania.^37,38

Neurotransmitter hypothesis. It has been shown that in patients with delirious mania there is dysregulation of dopamine transport, which leads to dopamine overflow in the synapse. In contrast to a drug effect (ie, cocaine or methamphetamine) that acts by inhibiting dopamine reuptake, dopamine overflow in delirious mania is caused by the loss of dopamine transporter regulation. This results in a dysfunctional dopaminergic state that precipitates an acute state of delirium and agitation.^39,40

Serotonin plays a role in mood disorders, including mania and depression.^41,42 More specifically, serotonin has been implicated in impulsivity and aggression as shown by reduced levels of CSF 5-hydroxyindoleacetic acid (5-HIAA) and depletion of 5-hydroxytryptophan (5-HTP).⁴³

Continue to: Alterations in gamma-aminobutyric acid (GABA) transmission...

Alterations in gamma-aminobutyric acid (GABA) transmission are known to occur in delirium and catatonia. In delirium, GABA signaling is increased, which disrupts the circadian rhythm and melatonin release, thus impairing the sleep-wake cycle.⁴⁴ Deficiencies in acetylcholine and melatonin are seen as well as excess of other neurotransmitters, including norepinephrine and glutamate.⁴⁵ Conversely, in catatonia, functional imaging studies found decreased GABA-A binding in orbito­frontal, prefrontal, parietal, and motor cortical regions.⁴⁶ A study analyzing 10 catatonic patients found decreased density of GABA-A receptors in the left sensorimotor cortex compared to psychiatric and healthy controls.⁴⁷

Other neurotransmitters, such as glutamate, at the N-methyl-D-aspartate receptors (NMDAR) have been hypothesized to be hyperactive, causing downstream dysregulation of GABA functioning.⁴⁸ However, the exact connection between delirious mania and all these receptors and neurotransmitters remains unknown.

Encephalitis hypothesis. The relationship between delirious mania and autoimmune encephalitis suggests delirious mania has etiologies other than a primary psychiatric illness. In a 2020 retrospective study⁴⁹ that analyzed 79 patients with anti-NMDAR encephalitis, 25.3% met criteria for delirious mania, and 95% of these patients had catatonic features. Dalmau et al⁵⁰ found that in many cases, patients tend to respond to ECT; in a cases series of 3 patients, 2 responded to benzodiazepines.

COVID-19 hypothesis. The SARS-CoV-2 virion has been associated with many neuropsychiatric complications, including mood, psychotic, and neurocognitive disorders.^51,52 There also have been cases of COVID-19–induced catatonia.^53-55 One case of delirious mania in a patient with COVID-19 has been reported.²¹ The general mechanism has been proposed to be related to the stimulation of the proinflammatory cytokines, such as tumor necrosis factor-alpha and interleukin-6, which the virus produces in large quantities.⁵⁶ These cytokines have been linked to psychosis and other psychiatric disorders.⁵⁷ The patient with COVID-19–induced delirious mania had elevated inflammatory markers, including erythrocyte sedimentation rate, C-reactive protein, ferritin, and D-dimer, which supports a proinflammatory state. This patient had a complete resolution of symptoms with ECT.²¹

Management: Benzodiazepines and ECT

A step-by-step algorithm for managing delirious mania is outlined in the Figure. Regardless of the underlining etiology, management of delirious mania consists of benzodiazepines (lorazepam and diazepam); prompt use of ECT, particularly for patients who do not improve with large doses of lorazepam; or (if applicable) continued treatment of the underlining medical condition, which does not preclude the use of benzodiazepines or ECT. Recent reports^27,58 have described details for using ECT for delirious mania, highlighting the use of high-energy dosing, bilateral electrode placement, and frequent sessions.

Continue to: Knowing which medications...

Knowing which medications to avoid is as important as knowing which agents to administer. Be vigilant in avoiding high-potency antipsychotics, as these medications can worsen extrapyramidal symptoms and may precipitate seizures or neuroleptic malignant syndrome (NMS).²⁸ Anticholinergic agents should also be avoided because they worsen confusion. Although lithium is effective in BD, in delirious mania, high doses of lithium and haloperidol may cause severe encephalopathic syndromes, with symptoms that can include lethargy, tremors, cerebellar dysfunction, and worsened confusion; it may also cause widespread and irreversible brain damage.⁵⁹While positive outcomes have been documented when using a combination of antipsychotics and lithium,^8,60 this approach should be considered carefully and tailored to individual cases, taking into account the severity of manic and psychotic symptoms in addition to the level of catatonia.

Due to long periods of hyperactivity, withdrawal, and diaphoresis, patients with delirious mania may be severely dehydrated with metabolic derangements, including elevated CPK due to rhabdomyolysis from prolonged exertion, IM antipsychotics, or rigidity. To prevent acute renal failure, this must be immediately addressed with rapid fluid resuscitation and electrolyte repletion.⁶¹

Benzodiazepines. The rapid use of lorazepam should be initiated when delirious mania is suspected. Doses of 6 to 20 mg have been reported to be effective if tolerated.^5,20 Typically, high-dose lorazepam will not have the sedative effect that would normally occur in a patient who does not have delirious mania.² Lorazepam should be titrated until full resolution of symptoms. Doses up to 30 mg have been reported as effective and tolerable.⁶² In our literature review, 50% of patients (18/36) responded or partially responded to lorazepam. However, only 3 case reports documented a complete remission with lorazepam, and many patients needed ECT for remission of symptoms.

ECT is generally reserved for patients who are not helped by benzodiazepine therapy, which is estimated to be up to 20%.⁵ ECT is highly effective in delirious mania, with remission rates ranging from 80% to 100%.¹ ECT is also effective in acute non­delirious mania (comparable to depression); however, it is only used in a small minority of cases (0.2% to 12%).³⁵ In our review, 58% of cases (21/36) reported using ECT, and in all cases it resulted in complete remission.

A dramatic improvement can be seen even after a single ECT session, though most patients show improvement after 4 sessions or 3 to 7 days.^1,2,5 In our review, most patients received 4 to 12 sessions until achieving complete remission.

Continue to: No RCTs have evaluated...

No RCTs have evaluated ECT electrode placement in patients with delirious mania. However, several RCTs have investigated electrode placement in patients with acute nondelirious mania. Hiremani et al⁶³ found that bitemporal placement had a more rapid response rate than bifrontal placement, but there was no overall difference in response rate. Barekatain et al⁶⁴ found no difference between these 2 bilateral approaches. Many of the delirious mania cases report using a bilateral placement (including 42% of the ECT cases in our review) due to the emergent need for rapid relief of symptoms, which is especially necessary if the patient is experiencing hemodynamic instability, excessive violence, risk for self-harm, worsening delirium, or resistance to lorazepam.

Prognosis: Often fatal if left untreated

Patients with delirious mania are at high risk to progress to a more severe form of NMS or malignant catatonia. Therefore, high-potency antipsychotics should be avoided; mortality can be elevated from 60% without antipsychotics to 78% with antipsychotics.⁴ Some researchers estimate 75% to 78% of cases of delirious mania can be fatal if left untreated.^3,6

Bottom Line

Delirious mania is routinely mistaken for more conventional manic or psychotic disorders. Clinicians need to be able to rapidly recognize the symptoms of this syndrome, which include mania, psychosis, delirium, and possible catatonia, so they can avoid administering toxic agents and instead initiate effective treatments such as benzodiazepines and electroconvulsive therapy.

Related Resources

• Arsan C, Baker C, Wong J, et al. Delirious mania: an approach to diagnosis and treatment. Prim Care Companion CNS Disord. 2021;23(1):20f02744. doi:10.4088/PCC.20f02744
• Lamba G, Kennedy EA, Vu CP. Case report: ECT for delirious mania. Clinical Psychiatry News. December 14, 2021. https://www.mdedge.com/psychiatry/article/249909/bipolar-disorder/case-report-ect-delirious-mania

Drug Brand Names

Diazepam • Valium
Haloperidol • Haldol
Lithium • Eskalith, Lithobid
Lorazepam • Ativan

Delirious mania is a syndrome characterized by the acute onset of severe hyperactivity, psychosis, catatonia, and intermittent confusion. While there have been growing reports of this phenomenon over the last 2 decades, it remains poorly recognized and understood.^1,2 There is no widely accepted nosology for delirious mania and the condition is absent from DSM-5, which magnifies the difficulties in making a timely diagnosis and initiating appropriate treatment. Delayed diagnosis and treatment may result in a detrimental outcome.^2,3 Delirious mania has also been labeled as lethal catatonia, specific febrile delirium, hyperactive or exhaustive mania, and Bell’s mania.^2,4,5 The characterization and diagnosis of this condition have a long and inconsistent history (Box^1,6-11).

Box

Delirious mania continues to be met with incertitude in clinical practice, and numerous inconsistencies have been reported in the literature. For example, some cases that have been reported as delirious mania had more evidence of primary delirium due to another medical condition or primary mania.^12,13 Other cases have demonstrated swift improvement of symptoms after monotherapy with antipsychotics without a trial of benzodiazepines or electroconvulsive therapy (ECT); the exclusion of a sudden onset questions the validity of the diagnosis and promotes the use of less efficacious treatments.^14,15 Other reports have confirmed that the diagnosis is missed when certain symptoms are more predominant, such as a thought disorder (acute schizophrenia), grandiosity and delusional ideation (bipolar disorder [BD]), and less commonly assessed catatonic signs (ambitendency, automatic obedience). These symptoms are mistakenly attributed to the respective disease.^1,16 This especially holds true when delirious mania is initially diagnosed as a primary psychosis, which leads to the administration of antipsychotics.¹⁷ Other cases have reported that delirious mania was resistant to treatment, but ECT was never pursued.¹⁸

In this review, we provide a more comprehensive perspective of the clinical presentation, pathogenesis, and management of delirious mania. We searched PubMed and Google Scholar using the keywords “delirious mania,” “delirious mania AND catatonia,” or “manic delirium.” Most articles we found were case reports, case series, or retrospective chart reviews. There were no systematic reviews, meta analyses, or randomized control trials (RCTs). The 12 articles included in this review consist of 7 individual case reports, 4 case series, and 1 retrospective chart review that describe a total of 36 cases (Table^{1,2,5,17,19-26}).

Clinical presentation: What to look for

Patients with delirious mania typically develop symptoms extremely rapidly. In virtually all published literature, symptoms were reported to emerge within hours to days and consisted of severe forms of mania, psychosis, and delirium; 100% of the cases in our review had these symptoms. Commonly reported symptoms were:

• intense excitement
• emotional lability
• grandiose delusions
• profound insomnia
• pressured and rapid speech
• auditory and visual hallucinations
• hypersexuality
• thought disorganization.

Exquisite paranoia can also result in violent aggression (and may require the use of physical restraints). Patients may confine themselves to very small spaces (such as a closet) in response to the intense paranoia. Impairments in various neurocognitive domains—including inability to focus; disorientation; language and visuospatial disturbances; difficulty with shifting and sustaining attention; and short-term memory impairments—have been reported. Patients often cannot recall the events during the episode.^1,2,5,27,28

Catatonia has been closely associated with delirious mania.²⁹ Features of excited catatonia—such as excessive motor activity, negativism, grimacing, posturing, echolalia, echopraxia, stereotypy, automatic obedience, verbigeration, combativeness, impulsivity, and rigidity—typically accompany delirious mania.^1,5,10,19,27

In addition to these symptoms, patients may engage in specific behaviors. They may exhibit inappropriate toileting such as smearing feces on walls or in bags, fecal or urinary incontinence, disrobing or running naked in public places, or pouring liquid on the floor or on one’s head.^1,2

Continue to: Of the 36 cases...

Of the 36 cases reported in the literature we reviewed, 20 (55%) were female. Most patients had an underlining psychiatric condition, including BD (72%), major depressive disorder (8%), and schizophrenia (2%). Three patients had no psychiatric history.

Physical examination

On initial presentation, a patient with delirious mania may be dehydrated, with dry mucous membranes, pale conjunctiva, tongue dryness, and poor skin turgor.^28,30 Due to excessive motor activity, diaphoresis with tachycardia, fluctuating blood pressure, and fever may be present.³¹

Certain basic cognitive tasks should be assessed to determine the patient’s orientation to place, date, and time. Assess if the patient can recall recent events, names of objects, or perform serial 7s; clock drawing capabilities also should be ascertained.^1,2,5 A Mini-Mental State Examination is useful.³²

The Bush-Francis Catatonia Rating Scale should be used to elicit features of catatonia, such as waxy flexibility, negativism, gegenhalten, mitgehen, catalepsy, ambitendency, automatic obedience, and grasp reflex.¹⁰

Laboratory findings are nonspecific

Although no specific laboratory findings are associated with delirious mania, bloodwork and imaging are routinely investigated, especially if delirium characteristics are most striking. A complete blood count, chemistries, hepatic panel, thyroid functioning, blood and/or urine cultures, creatinine phosphokinase (CPK), and urinalysis can be ordered. Head imaging such as MRI and CT to rule out intracranial pathology are typically performed.¹⁹ However, the diagnosis of delirious mania is based on the presence of the phenotypic features, by verification of catatonia, and by the responsiveness to the treatment delivered.²⁹

Continue to: Pathogenisis: Several hypotheses

The pathogenesis of delirious mania is not well understood. There are several postulations but no salient theory. Most patients with delirious mania have an underlying systemic medical or psychiatric condition.

Mood disorders. Patients with BD or schizoaffective disorder are especially susceptible to delirious mania. The percentage of manic patients who present with delirious mania varies by study. One study suggested approximately 19% have features of the phenomenon,³³ while others estimated 15% to 25%.³⁴ Elias et al³⁵ calculated that 15% of patients with mania succumb to manic exhaustion; from this it can be reasonably concluded that these were cases of misdiagnosed delirious mania.

Delirium hypothesis. Patients with delirious mania typically have features of delirium, including fluctuation of consciousness, disorientation, and/or poor sleep-wake cycle.³⁶ During rapid eye movement (REM) and non-REM sleep, memory circuits are fortified. When there is a substantial loss of REM and non-REM sleep, these circuits become faulty, even after 1 night. Pathological brain waves on EEG reflect the inability to reinforce the memory circuits. Patients with these waves may develop hallucinations, bizarre delusions, and altered sensorium. ECT reduces the pathological slow wave morphologies, thus restoring the synaptic maintenance and correcting the incompetent circuitry. This can explain the robust and rapid response of ECT in a patient with delirious mania.^37,38

Neurotransmitter hypothesis. It has been shown that in patients with delirious mania there is dysregulation of dopamine transport, which leads to dopamine overflow in the synapse. In contrast to a drug effect (ie, cocaine or methamphetamine) that acts by inhibiting dopamine reuptake, dopamine overflow in delirious mania is caused by the loss of dopamine transporter regulation. This results in a dysfunctional dopaminergic state that precipitates an acute state of delirium and agitation.^39,40

Serotonin plays a role in mood disorders, including mania and depression.^41,42 More specifically, serotonin has been implicated in impulsivity and aggression as shown by reduced levels of CSF 5-hydroxyindoleacetic acid (5-HIAA) and depletion of 5-hydroxytryptophan (5-HTP).⁴³

Continue to: Alterations in gamma-aminobutyric acid (GABA) transmission...

Alterations in gamma-aminobutyric acid (GABA) transmission are known to occur in delirium and catatonia. In delirium, GABA signaling is increased, which disrupts the circadian rhythm and melatonin release, thus impairing the sleep-wake cycle.⁴⁴ Deficiencies in acetylcholine and melatonin are seen as well as excess of other neurotransmitters, including norepinephrine and glutamate.⁴⁵ Conversely, in catatonia, functional imaging studies found decreased GABA-A binding in orbito­frontal, prefrontal, parietal, and motor cortical regions.⁴⁶ A study analyzing 10 catatonic patients found decreased density of GABA-A receptors in the left sensorimotor cortex compared to psychiatric and healthy controls.⁴⁷

Other neurotransmitters, such as glutamate, at the N-methyl-D-aspartate receptors (NMDAR) have been hypothesized to be hyperactive, causing downstream dysregulation of GABA functioning.⁴⁸ However, the exact connection between delirious mania and all these receptors and neurotransmitters remains unknown.

Encephalitis hypothesis. The relationship between delirious mania and autoimmune encephalitis suggests delirious mania has etiologies other than a primary psychiatric illness. In a 2020 retrospective study⁴⁹ that analyzed 79 patients with anti-NMDAR encephalitis, 25.3% met criteria for delirious mania, and 95% of these patients had catatonic features. Dalmau et al⁵⁰ found that in many cases, patients tend to respond to ECT; in a cases series of 3 patients, 2 responded to benzodiazepines.

COVID-19 hypothesis. The SARS-CoV-2 virion has been associated with many neuropsychiatric complications, including mood, psychotic, and neurocognitive disorders.^51,52 There also have been cases of COVID-19–induced catatonia.^53-55 One case of delirious mania in a patient with COVID-19 has been reported.²¹ The general mechanism has been proposed to be related to the stimulation of the proinflammatory cytokines, such as tumor necrosis factor-alpha and interleukin-6, which the virus produces in large quantities.⁵⁶ These cytokines have been linked to psychosis and other psychiatric disorders.⁵⁷ The patient with COVID-19–induced delirious mania had elevated inflammatory markers, including erythrocyte sedimentation rate, C-reactive protein, ferritin, and D-dimer, which supports a proinflammatory state. This patient had a complete resolution of symptoms with ECT.²¹

Management: Benzodiazepines and ECT

A step-by-step algorithm for managing delirious mania is outlined in the Figure. Regardless of the underlining etiology, management of delirious mania consists of benzodiazepines (lorazepam and diazepam); prompt use of ECT, particularly for patients who do not improve with large doses of lorazepam; or (if applicable) continued treatment of the underlining medical condition, which does not preclude the use of benzodiazepines or ECT. Recent reports^27,58 have described details for using ECT for delirious mania, highlighting the use of high-energy dosing, bilateral electrode placement, and frequent sessions.

Continue to: Knowing which medications...

Knowing which medications to avoid is as important as knowing which agents to administer. Be vigilant in avoiding high-potency antipsychotics, as these medications can worsen extrapyramidal symptoms and may precipitate seizures or neuroleptic malignant syndrome (NMS).²⁸ Anticholinergic agents should also be avoided because they worsen confusion. Although lithium is effective in BD, in delirious mania, high doses of lithium and haloperidol may cause severe encephalopathic syndromes, with symptoms that can include lethargy, tremors, cerebellar dysfunction, and worsened confusion; it may also cause widespread and irreversible brain damage.⁵⁹While positive outcomes have been documented when using a combination of antipsychotics and lithium,^8,60 this approach should be considered carefully and tailored to individual cases, taking into account the severity of manic and psychotic symptoms in addition to the level of catatonia.

Due to long periods of hyperactivity, withdrawal, and diaphoresis, patients with delirious mania may be severely dehydrated with metabolic derangements, including elevated CPK due to rhabdomyolysis from prolonged exertion, IM antipsychotics, or rigidity. To prevent acute renal failure, this must be immediately addressed with rapid fluid resuscitation and electrolyte repletion.⁶¹

Benzodiazepines. The rapid use of lorazepam should be initiated when delirious mania is suspected. Doses of 6 to 20 mg have been reported to be effective if tolerated.^5,20 Typically, high-dose lorazepam will not have the sedative effect that would normally occur in a patient who does not have delirious mania.² Lorazepam should be titrated until full resolution of symptoms. Doses up to 30 mg have been reported as effective and tolerable.⁶² In our literature review, 50% of patients (18/36) responded or partially responded to lorazepam. However, only 3 case reports documented a complete remission with lorazepam, and many patients needed ECT for remission of symptoms.

ECT is generally reserved for patients who are not helped by benzodiazepine therapy, which is estimated to be up to 20%.⁵ ECT is highly effective in delirious mania, with remission rates ranging from 80% to 100%.¹ ECT is also effective in acute non­delirious mania (comparable to depression); however, it is only used in a small minority of cases (0.2% to 12%).³⁵ In our review, 58% of cases (21/36) reported using ECT, and in all cases it resulted in complete remission.

A dramatic improvement can be seen even after a single ECT session, though most patients show improvement after 4 sessions or 3 to 7 days.^1,2,5 In our review, most patients received 4 to 12 sessions until achieving complete remission.

Continue to: No RCTs have evaluated...

No RCTs have evaluated ECT electrode placement in patients with delirious mania. However, several RCTs have investigated electrode placement in patients with acute nondelirious mania. Hiremani et al⁶³ found that bitemporal placement had a more rapid response rate than bifrontal placement, but there was no overall difference in response rate. Barekatain et al⁶⁴ found no difference between these 2 bilateral approaches. Many of the delirious mania cases report using a bilateral placement (including 42% of the ECT cases in our review) due to the emergent need for rapid relief of symptoms, which is especially necessary if the patient is experiencing hemodynamic instability, excessive violence, risk for self-harm, worsening delirium, or resistance to lorazepam.

Prognosis: Often fatal if left untreated

Patients with delirious mania are at high risk to progress to a more severe form of NMS or malignant catatonia. Therefore, high-potency antipsychotics should be avoided; mortality can be elevated from 60% without antipsychotics to 78% with antipsychotics.⁴ Some researchers estimate 75% to 78% of cases of delirious mania can be fatal if left untreated.^3,6

Bottom Line

Delirious mania is routinely mistaken for more conventional manic or psychotic disorders. Clinicians need to be able to rapidly recognize the symptoms of this syndrome, which include mania, psychosis, delirium, and possible catatonia, so they can avoid administering toxic agents and instead initiate effective treatments such as benzodiazepines and electroconvulsive therapy.

Related Resources

• Arsan C, Baker C, Wong J, et al. Delirious mania: an approach to diagnosis and treatment. Prim Care Companion CNS Disord. 2021;23(1):20f02744. doi:10.4088/PCC.20f02744
• Lamba G, Kennedy EA, Vu CP. Case report: ECT for delirious mania. Clinical Psychiatry News. December 14, 2021. https://www.mdedge.com/psychiatry/article/249909/bipolar-disorder/case-report-ect-delirious-mania

Drug Brand Names

Diazepam • Valium
Haloperidol • Haldol
Lithium • Eskalith, Lithobid
Lorazepam • Ativan

After several months of difficulty living in her current apartment complex, Ms. M asks you as her psychiatrist to write a letter to the management company requesting she be moved to an apartment on the opposite side of the maintenance closet because the noise aggravates her posttraumatic stress disorder. What should you consider when asked to write such a letter?

Psychiatric practice often extends beyond the treatment of mental illness to include addressing patients’ social well-being. Psychiatrists commonly inquire about a patient’s social situation to understand the impact of these environmental factors. Similarly, psychiatric illness may affect a patient’s ability to work or fulfill responsibilities. As a result, patients may ask their psychiatrists for assistance by requesting letters that address various aspects of their social well-being.¹ These communications may address an array of topics, from a patient’s readiness to return to work to their ability to pay child support. This article focuses on the role psychiatrists have in writing patient-requested letters across a variety of topics, including the consideration of potential legal liability and ethical implications.

Types of letters

The categories of letters patients request can be divided into 2 groups. The first is comprised of letters relating to the patient’s medical needs (Table 1^2,3). These address the patient’s ability to work (eg, medical leave, return to work, or accommodations) or travel (eg, ability to drive or use public transportation), or need for specific medical treatment (ie, gender-affirming care or cannabis use in specific settings). The second group relates to legal requests such as excusal from jury duty, emotional support animals, or any other letter used specifically for legal purposes (in civil or criminal cases) (Table 2^1,4-6).

The decision to write a letter on behalf of a patient should be based on whether you have sufficient knowledge to answer the referral question, and whether the requested evaluation fits within your role as the treating psychiatrist. Many requests fall short of the first condition. For example, a request to opine about an individual’s ability to perform their job duties requires specific knowledge and careful consideration of the patient’s work responsibilities, knowledge of the impact of their psychiatric symptoms, and specialized knowledge about interventions that would ameliorate symptoms in the specialized work setting. Most psychiatrists are not sufficiently familiar with a specific workplace to provide opinions regarding reasonable accommodations.

The second condition refers to the role and responsibilities of the psychiatrist. Many letter requests are clearly within the scope of the clinical psychiatrist, such as a medical leave note due to a psychiatric decompensation or a jury duty excusal due to an unstable mental state. Other letters reach beyond the role of the general or treating psychiatrist, such as opinions about suitable housing or a patient’s competency to stand trial.

Components of letters

The decision to write or not to write a letter should be discussed with the patient. Identify the reasons for and against letter writing. If you decide to write a letter, the letter should have the following basic framework (Figure): the identity of the person who requested the letter, the referral question, and an answer to the referral question with a clear rationale. Describe the patient’s psychiatric diagnosis using DSM criteria. Any limitations to the answer should be identified. The letter should not go beyond the referral question and should not include information that was not requested. It also should be preserved in the medical record.

It is recommended to write or review the letter in the presence of the patient to discuss the contents of the letter and what the psychiatrist can or cannot write. As in forensic reports, conclusory statements are not helpful. Provide descriptive information instead of relying on psychiatric jargon, and a rationale for the opinion as opposed to stating an opinion as fact. In the letter, you must acknowledge that your opinion is based upon information provided by the patient (and the patient’s family, when accurate) and as a result, is not fully objective.

Continue to: Liability and dual agency

Psychiatrists are familiar with clinical situations in which a duty to the patient is mitigated or superseded by a duty to a third party. As the Tarasoff court famously stated, “the protective privilege ends where the public peril begins.”⁷

To be liable to either a patient or a third party means to be “bound or obliged in law or equity; responsible; chargeable; answerable; compellable to make satisfaction, compensation, or restitution.”⁸ Liabilities related to clinical treatment are well-established; medical students learn the fundamentals before ever treating a patient, and physicians carry malpractice insurance throughout their careers.

Less well-established is the liability a treating psychiatrist owes a third party when forming an opinion that impacts both their patient and the third party (eg, an employer when writing a return-to-work letter, or a disability insurer when qualifying a patient for disability benefits). The American Academy of Psychiatry and the Law discourages treating psychiatrists from performing these types of evaluations of their patients based on the inherent conflict of serving as a dual agent, or acting both as an advocate for the patient and as an independent evaluator striving for objectivity.⁹ However, such requests commonly arise, and some may be unavoidable.

Dual-agency situations subject the treating psychiatrist to avenues of legal action arising from the patient-doctor relationship as well as the forensic evaluator relationship. If a letter is written during a clinical treatment, all duties owed to the patient continue to apply, and the relevant benchmarks of local statutes and principle of a standard of care are relevant. It is conceivable that a patient could bring a negligence lawsuit based on a standard of care allegation (eg, that writing certain types of letters is so ordinary that failure to write them would fall below the standard of care). Confidentiality is also of the utmost importance,¹⁰ and you should obtain a written release of information from the patient before releasing any letter with privileged information about the patient.¹¹ Additional relevant legal causes of action the patient could include are torts such as defamation of character, invasion of privacy, breach of contract, and intentional infliction of emotional distress. There is limited case law supporting patients’ rights to sue psychiatrists for defamation.¹⁰

A psychiatrist writing a letter to a third party may also subject themselves to avenues of legal action occurring outside the physician-patient relationship. Importantly, damages resulting from these breaches would not be covered by your malpractice insurance. Extreme cases involve allegations of fraud or perjury, which could be pursued in criminal court. If a psychiatrist intentionally deceives a third party for the purpose of obtaining some benefit for the patient, this is clear grounds for civil or criminal action. Fraud is defined as “a false representation of a matter of fact, whether by words or by conduct, by false or misleading allegations, or by concealment of that which should have been disclosed, which deceives and is intended to deceive another so that he shall act upon it to his legal injury.”⁸ Negligence can also be grounds for liability if a third party suffers injury or loss. Although the liability is clearer if the third party retains an independent psychiatrist rather than soliciting an opinion from a patient’s treating psychiatrist, both parties are subject to the claim of negligence.¹⁰

Continue to: There are some important protections...

There are some important protections that limit psychiatrists’ good-faith opinions from litigation. The primary one is the “professional medical judgment rule,” which shields physicians from the consequences of erroneous opinions so long as the examination was competent, complete, and performed in an ordinary fashion.¹⁰ In some cases, psychiatrists writing a letter or report for a government agency may also qualify for quasi-judicial immunity or witness immunity, but case law shows significant variation in when and how these privileges apply and whether such privileges would be applied to a clinical psychiatrist in the context of a traditional physician-patient relationship.¹² In general, these privileges are not absolute and may not be sufficiently well-established to discourage a plaintiff from filing suit or prompt early judicial dismissal of a case.

Like all aspects of practicing medicine, letter writing is subject to scrutiny and accountability. Think carefully about your obligations and the potential consequences of writing or not writing a letter to a third party.

Ethical considerations

The decision to write a letter for a patient must be carefully considered from multiple angles.⁶ In addition to liability concerns, various ethical considerations also arise. Guided by the principles of beneficence, nonmaleficence, autonomy, and justice,¹³ we recommend the following approaches.

Maintain objectivity

During letter writing, a conflict of interest may arise between your allegiance to the patient and the imperative to provide accurate information.^14-16 If the conflict is overwhelming, the most appropriate approach is to recuse yourself from the case and refer the patient to a third party. When electing to write a letter, you accept the responsibility to provide an objective assessment of the relevant situation. This promotes a just outcome and may also serve to promote the patient’s or society’s well-being.

Encourage activity and overall function

Evidence suggests that participation in multiple aspects of life promotes positive health outcomes.^17,18 As a physician, it is your duty to promote health and support and facilitate accommodations that allow patients to participate and flourish in society. By the same logic, when approached by patients with a request for letters in support of reduced activity, you should consider not only the benefits but also the potential detriments of such disruptions. This may entail recommending temporary restrictions or modifications, as appropriate.

Continue to: Think beyond the patient

Letter writing, particularly when recommending accommodations, can have implications beyond the patient.¹⁶ Such letters may cause unintended societal harm. For example, others may have to assume additional responsibilities; competitive goods (eg, housing) may be rendered to the patient rather than to a person with greater needs; and workplace safety could be compromised due to absence. Consider not only the individual patient but also possible public health and societal effects of letter writing.

Deciding not to write

From an ethical perspective, a physician cannot be compelled to write a letter if such an undertaking violates a stronger moral obligation. An example of this is if writing a letter could cause significant harm to the patient or society, or if writing a letter might compromise a physician’s professionalism.¹⁹ When you elect to not write a letter, the ethical principles of autonomy and truth telling dictate that you must inform your patients of this choice.⁶ You should also provide an explanation to the patient as long as such information would not cause undue psychological or physical harm.^20,21

Schedule time to write letters

Some physicians implement policies that all letters are to be completed during scheduled appointments. Others designate administrative time to complete requested letters. Finally, some physicians flexibly complete such requests between appointments or during other undedicated time slots. Any of these approaches are justifiable, though some urgent requests may require more immediate attention outside of appointments. Some physicians may choose to bill for the letter writing if completed outside an appointment and the patient is treated in private practice. Whatever your policy, inform patients of it at the beginning of care and remind them when appropriate, such as before completing a letter that may be billed.

Manage uncertainty

Always strive for objectivity in letter writing. However, some requests inherently hinge on subjective reports and assessments. For example, a patient may request an excuse letter due to feeling unwell. In the absence of objective findings, what should you do? We advise the following.

Acquire collateral information. Adequate information is essential when making any medical recommendation. The same is true for writing letters. With the patient’s permission, you may need to contact relevant parties to better understand the circumstance or activity about which you are being asked to write a letter. For example, a patient may request leave from work due to injury. If the specific parameters of the work impeded by the injury are unclear to you, refrain from writing the letter and explain the rationale to the patient.

Continue to: Integrate prior knowledge of the patient

Integrate prior knowledge of the patient. No letter writing request exists in a vacuum. If you know the patient, the letter should be contextualized within the patient’s prior behaviors.

Stay within your scope

Given the various dilemmas and challenges, you may want to consider whether some letter writing is out of your professional scope.^14-16 One solution would be to leave such requests to other entities (eg, requiring employers to retain medical personnel with specialized skills in occupational evaluations) and make such recommendations to patients. Regardless, physicians should think carefully about their professional boundaries and scope regarding letter requests and adopt and implement a consistent standard for all patients.

Regarding the letter requested by Ms. M, you should consider whether the appeal is consistent with the patient’s psychiatric illness. You should also consider whether you have sufficient knowledge about the patient’s living environment to support their claim. Such a letter should be written only if you understand both considerations. Regardless of your decision, you should explain your rationale to the patient.

Bottom Line

Patients may ask their psychiatrists to write letters that address aspects of their social well-being. However, psychiatrists must be alert to requests that are outside their scope of practice or ethically or legally fraught. Carefully consider whether writing a letter is appropriate and if not, discuss with the patient the reasons you cannot write such a letter and any recommended alternative avenues to address their request.

Related Resources

• Riese A. Writing letters for transgender patients undergoing medical transition. Current Psychiatry. 2021;20(8):51-52. doi:10.12788/cp.0159
• Joshi KG. Service animals and emotional support animals: should you write that letter? Current Psychiatry. 2021;20(11):16-19,24. doi:10.12788/cp.0183

After several months of difficulty living in her current apartment complex, Ms. M asks you as her psychiatrist to write a letter to the management company requesting she be moved to an apartment on the opposite side of the maintenance closet because the noise aggravates her posttraumatic stress disorder. What should you consider when asked to write such a letter?

Psychiatric practice often extends beyond the treatment of mental illness to include addressing patients’ social well-being. Psychiatrists commonly inquire about a patient’s social situation to understand the impact of these environmental factors. Similarly, psychiatric illness may affect a patient’s ability to work or fulfill responsibilities. As a result, patients may ask their psychiatrists for assistance by requesting letters that address various aspects of their social well-being.¹ These communications may address an array of topics, from a patient’s readiness to return to work to their ability to pay child support. This article focuses on the role psychiatrists have in writing patient-requested letters across a variety of topics, including the consideration of potential legal liability and ethical implications.

Types of letters

The categories of letters patients request can be divided into 2 groups. The first is comprised of letters relating to the patient’s medical needs (Table 1^2,3). These address the patient’s ability to work (eg, medical leave, return to work, or accommodations) or travel (eg, ability to drive or use public transportation), or need for specific medical treatment (ie, gender-affirming care or cannabis use in specific settings). The second group relates to legal requests such as excusal from jury duty, emotional support animals, or any other letter used specifically for legal purposes (in civil or criminal cases) (Table 2^1,4-6).

The decision to write a letter on behalf of a patient should be based on whether you have sufficient knowledge to answer the referral question, and whether the requested evaluation fits within your role as the treating psychiatrist. Many requests fall short of the first condition. For example, a request to opine about an individual’s ability to perform their job duties requires specific knowledge and careful consideration of the patient’s work responsibilities, knowledge of the impact of their psychiatric symptoms, and specialized knowledge about interventions that would ameliorate symptoms in the specialized work setting. Most psychiatrists are not sufficiently familiar with a specific workplace to provide opinions regarding reasonable accommodations.

The second condition refers to the role and responsibilities of the psychiatrist. Many letter requests are clearly within the scope of the clinical psychiatrist, such as a medical leave note due to a psychiatric decompensation or a jury duty excusal due to an unstable mental state. Other letters reach beyond the role of the general or treating psychiatrist, such as opinions about suitable housing or a patient’s competency to stand trial.

Components of letters

The decision to write or not to write a letter should be discussed with the patient. Identify the reasons for and against letter writing. If you decide to write a letter, the letter should have the following basic framework (Figure): the identity of the person who requested the letter, the referral question, and an answer to the referral question with a clear rationale. Describe the patient’s psychiatric diagnosis using DSM criteria. Any limitations to the answer should be identified. The letter should not go beyond the referral question and should not include information that was not requested. It also should be preserved in the medical record.

It is recommended to write or review the letter in the presence of the patient to discuss the contents of the letter and what the psychiatrist can or cannot write. As in forensic reports, conclusory statements are not helpful. Provide descriptive information instead of relying on psychiatric jargon, and a rationale for the opinion as opposed to stating an opinion as fact. In the letter, you must acknowledge that your opinion is based upon information provided by the patient (and the patient’s family, when accurate) and as a result, is not fully objective.

Continue to: Liability and dual agency

Psychiatrists are familiar with clinical situations in which a duty to the patient is mitigated or superseded by a duty to a third party. As the Tarasoff court famously stated, “the protective privilege ends where the public peril begins.”⁷

To be liable to either a patient or a third party means to be “bound or obliged in law or equity; responsible; chargeable; answerable; compellable to make satisfaction, compensation, or restitution.”⁸ Liabilities related to clinical treatment are well-established; medical students learn the fundamentals before ever treating a patient, and physicians carry malpractice insurance throughout their careers.

Less well-established is the liability a treating psychiatrist owes a third party when forming an opinion that impacts both their patient and the third party (eg, an employer when writing a return-to-work letter, or a disability insurer when qualifying a patient for disability benefits). The American Academy of Psychiatry and the Law discourages treating psychiatrists from performing these types of evaluations of their patients based on the inherent conflict of serving as a dual agent, or acting both as an advocate for the patient and as an independent evaluator striving for objectivity.⁹ However, such requests commonly arise, and some may be unavoidable.

Dual-agency situations subject the treating psychiatrist to avenues of legal action arising from the patient-doctor relationship as well as the forensic evaluator relationship. If a letter is written during a clinical treatment, all duties owed to the patient continue to apply, and the relevant benchmarks of local statutes and principle of a standard of care are relevant. It is conceivable that a patient could bring a negligence lawsuit based on a standard of care allegation (eg, that writing certain types of letters is so ordinary that failure to write them would fall below the standard of care). Confidentiality is also of the utmost importance,¹⁰ and you should obtain a written release of information from the patient before releasing any letter with privileged information about the patient.¹¹ Additional relevant legal causes of action the patient could include are torts such as defamation of character, invasion of privacy, breach of contract, and intentional infliction of emotional distress. There is limited case law supporting patients’ rights to sue psychiatrists for defamation.¹⁰

A psychiatrist writing a letter to a third party may also subject themselves to avenues of legal action occurring outside the physician-patient relationship. Importantly, damages resulting from these breaches would not be covered by your malpractice insurance. Extreme cases involve allegations of fraud or perjury, which could be pursued in criminal court. If a psychiatrist intentionally deceives a third party for the purpose of obtaining some benefit for the patient, this is clear grounds for civil or criminal action. Fraud is defined as “a false representation of a matter of fact, whether by words or by conduct, by false or misleading allegations, or by concealment of that which should have been disclosed, which deceives and is intended to deceive another so that he shall act upon it to his legal injury.”⁸ Negligence can also be grounds for liability if a third party suffers injury or loss. Although the liability is clearer if the third party retains an independent psychiatrist rather than soliciting an opinion from a patient’s treating psychiatrist, both parties are subject to the claim of negligence.¹⁰

Continue to: There are some important protections...

There are some important protections that limit psychiatrists’ good-faith opinions from litigation. The primary one is the “professional medical judgment rule,” which shields physicians from the consequences of erroneous opinions so long as the examination was competent, complete, and performed in an ordinary fashion.¹⁰ In some cases, psychiatrists writing a letter or report for a government agency may also qualify for quasi-judicial immunity or witness immunity, but case law shows significant variation in when and how these privileges apply and whether such privileges would be applied to a clinical psychiatrist in the context of a traditional physician-patient relationship.¹² In general, these privileges are not absolute and may not be sufficiently well-established to discourage a plaintiff from filing suit or prompt early judicial dismissal of a case.

Like all aspects of practicing medicine, letter writing is subject to scrutiny and accountability. Think carefully about your obligations and the potential consequences of writing or not writing a letter to a third party.

Ethical considerations

The decision to write a letter for a patient must be carefully considered from multiple angles.⁶ In addition to liability concerns, various ethical considerations also arise. Guided by the principles of beneficence, nonmaleficence, autonomy, and justice,¹³ we recommend the following approaches.

Maintain objectivity

During letter writing, a conflict of interest may arise between your allegiance to the patient and the imperative to provide accurate information.^14-16 If the conflict is overwhelming, the most appropriate approach is to recuse yourself from the case and refer the patient to a third party. When electing to write a letter, you accept the responsibility to provide an objective assessment of the relevant situation. This promotes a just outcome and may also serve to promote the patient’s or society’s well-being.

Encourage activity and overall function

Evidence suggests that participation in multiple aspects of life promotes positive health outcomes.^17,18 As a physician, it is your duty to promote health and support and facilitate accommodations that allow patients to participate and flourish in society. By the same logic, when approached by patients with a request for letters in support of reduced activity, you should consider not only the benefits but also the potential detriments of such disruptions. This may entail recommending temporary restrictions or modifications, as appropriate.

Continue to: Think beyond the patient

Letter writing, particularly when recommending accommodations, can have implications beyond the patient.¹⁶ Such letters may cause unintended societal harm. For example, others may have to assume additional responsibilities; competitive goods (eg, housing) may be rendered to the patient rather than to a person with greater needs; and workplace safety could be compromised due to absence. Consider not only the individual patient but also possible public health and societal effects of letter writing.

Deciding not to write

From an ethical perspective, a physician cannot be compelled to write a letter if such an undertaking violates a stronger moral obligation. An example of this is if writing a letter could cause significant harm to the patient or society, or if writing a letter might compromise a physician’s professionalism.¹⁹ When you elect to not write a letter, the ethical principles of autonomy and truth telling dictate that you must inform your patients of this choice.⁶ You should also provide an explanation to the patient as long as such information would not cause undue psychological or physical harm.^20,21

Schedule time to write letters

Some physicians implement policies that all letters are to be completed during scheduled appointments. Others designate administrative time to complete requested letters. Finally, some physicians flexibly complete such requests between appointments or during other undedicated time slots. Any of these approaches are justifiable, though some urgent requests may require more immediate attention outside of appointments. Some physicians may choose to bill for the letter writing if completed outside an appointment and the patient is treated in private practice. Whatever your policy, inform patients of it at the beginning of care and remind them when appropriate, such as before completing a letter that may be billed.

Manage uncertainty

Always strive for objectivity in letter writing. However, some requests inherently hinge on subjective reports and assessments. For example, a patient may request an excuse letter due to feeling unwell. In the absence of objective findings, what should you do? We advise the following.

Acquire collateral information. Adequate information is essential when making any medical recommendation. The same is true for writing letters. With the patient’s permission, you may need to contact relevant parties to better understand the circumstance or activity about which you are being asked to write a letter. For example, a patient may request leave from work due to injury. If the specific parameters of the work impeded by the injury are unclear to you, refrain from writing the letter and explain the rationale to the patient.

Continue to: Integrate prior knowledge of the patient

Integrate prior knowledge of the patient. No letter writing request exists in a vacuum. If you know the patient, the letter should be contextualized within the patient’s prior behaviors.

Stay within your scope

Given the various dilemmas and challenges, you may want to consider whether some letter writing is out of your professional scope.^14-16 One solution would be to leave such requests to other entities (eg, requiring employers to retain medical personnel with specialized skills in occupational evaluations) and make such recommendations to patients. Regardless, physicians should think carefully about their professional boundaries and scope regarding letter requests and adopt and implement a consistent standard for all patients.

Regarding the letter requested by Ms. M, you should consider whether the appeal is consistent with the patient’s psychiatric illness. You should also consider whether you have sufficient knowledge about the patient’s living environment to support their claim. Such a letter should be written only if you understand both considerations. Regardless of your decision, you should explain your rationale to the patient.

Bottom Line

Patients may ask their psychiatrists to write letters that address aspects of their social well-being. However, psychiatrists must be alert to requests that are outside their scope of practice or ethically or legally fraught. Carefully consider whether writing a letter is appropriate and if not, discuss with the patient the reasons you cannot write such a letter and any recommended alternative avenues to address their request.

Related Resources

• Riese A. Writing letters for transgender patients undergoing medical transition. Current Psychiatry. 2021;20(8):51-52. doi:10.12788/cp.0159
• Joshi KG. Service animals and emotional support animals: should you write that letter? Current Psychiatry. 2021;20(11):16-19,24. doi:10.12788/cp.0183

After several months of difficulty living in her current apartment complex, Ms. M asks you as her psychiatrist to write a letter to the management company requesting she be moved to an apartment on the opposite side of the maintenance closet because the noise aggravates her posttraumatic stress disorder. What should you consider when asked to write such a letter?

Psychiatric practice often extends beyond the treatment of mental illness to include addressing patients’ social well-being. Psychiatrists commonly inquire about a patient’s social situation to understand the impact of these environmental factors. Similarly, psychiatric illness may affect a patient’s ability to work or fulfill responsibilities. As a result, patients may ask their psychiatrists for assistance by requesting letters that address various aspects of their social well-being.¹ These communications may address an array of topics, from a patient’s readiness to return to work to their ability to pay child support. This article focuses on the role psychiatrists have in writing patient-requested letters across a variety of topics, including the consideration of potential legal liability and ethical implications.

Types of letters

The categories of letters patients request can be divided into 2 groups. The first is comprised of letters relating to the patient’s medical needs (Table 1^2,3). These address the patient’s ability to work (eg, medical leave, return to work, or accommodations) or travel (eg, ability to drive or use public transportation), or need for specific medical treatment (ie, gender-affirming care or cannabis use in specific settings). The second group relates to legal requests such as excusal from jury duty, emotional support animals, or any other letter used specifically for legal purposes (in civil or criminal cases) (Table 2^1,4-6).

The decision to write a letter on behalf of a patient should be based on whether you have sufficient knowledge to answer the referral question, and whether the requested evaluation fits within your role as the treating psychiatrist. Many requests fall short of the first condition. For example, a request to opine about an individual’s ability to perform their job duties requires specific knowledge and careful consideration of the patient’s work responsibilities, knowledge of the impact of their psychiatric symptoms, and specialized knowledge about interventions that would ameliorate symptoms in the specialized work setting. Most psychiatrists are not sufficiently familiar with a specific workplace to provide opinions regarding reasonable accommodations.

The second condition refers to the role and responsibilities of the psychiatrist. Many letter requests are clearly within the scope of the clinical psychiatrist, such as a medical leave note due to a psychiatric decompensation or a jury duty excusal due to an unstable mental state. Other letters reach beyond the role of the general or treating psychiatrist, such as opinions about suitable housing or a patient’s competency to stand trial.

Components of letters

The decision to write or not to write a letter should be discussed with the patient. Identify the reasons for and against letter writing. If you decide to write a letter, the letter should have the following basic framework (Figure): the identity of the person who requested the letter, the referral question, and an answer to the referral question with a clear rationale. Describe the patient’s psychiatric diagnosis using DSM criteria. Any limitations to the answer should be identified. The letter should not go beyond the referral question and should not include information that was not requested. It also should be preserved in the medical record.

It is recommended to write or review the letter in the presence of the patient to discuss the contents of the letter and what the psychiatrist can or cannot write. As in forensic reports, conclusory statements are not helpful. Provide descriptive information instead of relying on psychiatric jargon, and a rationale for the opinion as opposed to stating an opinion as fact. In the letter, you must acknowledge that your opinion is based upon information provided by the patient (and the patient’s family, when accurate) and as a result, is not fully objective.

Continue to: Liability and dual agency

Psychiatrists are familiar with clinical situations in which a duty to the patient is mitigated or superseded by a duty to a third party. As the Tarasoff court famously stated, “the protective privilege ends where the public peril begins.”⁷

To be liable to either a patient or a third party means to be “bound or obliged in law or equity; responsible; chargeable; answerable; compellable to make satisfaction, compensation, or restitution.”⁸ Liabilities related to clinical treatment are well-established; medical students learn the fundamentals before ever treating a patient, and physicians carry malpractice insurance throughout their careers.

Less well-established is the liability a treating psychiatrist owes a third party when forming an opinion that impacts both their patient and the third party (eg, an employer when writing a return-to-work letter, or a disability insurer when qualifying a patient for disability benefits). The American Academy of Psychiatry and the Law discourages treating psychiatrists from performing these types of evaluations of their patients based on the inherent conflict of serving as a dual agent, or acting both as an advocate for the patient and as an independent evaluator striving for objectivity.⁹ However, such requests commonly arise, and some may be unavoidable.

Dual-agency situations subject the treating psychiatrist to avenues of legal action arising from the patient-doctor relationship as well as the forensic evaluator relationship. If a letter is written during a clinical treatment, all duties owed to the patient continue to apply, and the relevant benchmarks of local statutes and principle of a standard of care are relevant. It is conceivable that a patient could bring a negligence lawsuit based on a standard of care allegation (eg, that writing certain types of letters is so ordinary that failure to write them would fall below the standard of care). Confidentiality is also of the utmost importance,¹⁰ and you should obtain a written release of information from the patient before releasing any letter with privileged information about the patient.¹¹ Additional relevant legal causes of action the patient could include are torts such as defamation of character, invasion of privacy, breach of contract, and intentional infliction of emotional distress. There is limited case law supporting patients’ rights to sue psychiatrists for defamation.¹⁰

A psychiatrist writing a letter to a third party may also subject themselves to avenues of legal action occurring outside the physician-patient relationship. Importantly, damages resulting from these breaches would not be covered by your malpractice insurance. Extreme cases involve allegations of fraud or perjury, which could be pursued in criminal court. If a psychiatrist intentionally deceives a third party for the purpose of obtaining some benefit for the patient, this is clear grounds for civil or criminal action. Fraud is defined as “a false representation of a matter of fact, whether by words or by conduct, by false or misleading allegations, or by concealment of that which should have been disclosed, which deceives and is intended to deceive another so that he shall act upon it to his legal injury.”⁸ Negligence can also be grounds for liability if a third party suffers injury or loss. Although the liability is clearer if the third party retains an independent psychiatrist rather than soliciting an opinion from a patient’s treating psychiatrist, both parties are subject to the claim of negligence.¹⁰

Continue to: There are some important protections...

There are some important protections that limit psychiatrists’ good-faith opinions from litigation. The primary one is the “professional medical judgment rule,” which shields physicians from the consequences of erroneous opinions so long as the examination was competent, complete, and performed in an ordinary fashion.¹⁰ In some cases, psychiatrists writing a letter or report for a government agency may also qualify for quasi-judicial immunity or witness immunity, but case law shows significant variation in when and how these privileges apply and whether such privileges would be applied to a clinical psychiatrist in the context of a traditional physician-patient relationship.¹² In general, these privileges are not absolute and may not be sufficiently well-established to discourage a plaintiff from filing suit or prompt early judicial dismissal of a case.

Like all aspects of practicing medicine, letter writing is subject to scrutiny and accountability. Think carefully about your obligations and the potential consequences of writing or not writing a letter to a third party.

Ethical considerations

The decision to write a letter for a patient must be carefully considered from multiple angles.⁶ In addition to liability concerns, various ethical considerations also arise. Guided by the principles of beneficence, nonmaleficence, autonomy, and justice,¹³ we recommend the following approaches.

Maintain objectivity

During letter writing, a conflict of interest may arise between your allegiance to the patient and the imperative to provide accurate information.^14-16 If the conflict is overwhelming, the most appropriate approach is to recuse yourself from the case and refer the patient to a third party. When electing to write a letter, you accept the responsibility to provide an objective assessment of the relevant situation. This promotes a just outcome and may also serve to promote the patient’s or society’s well-being.

Encourage activity and overall function

Evidence suggests that participation in multiple aspects of life promotes positive health outcomes.^17,18 As a physician, it is your duty to promote health and support and facilitate accommodations that allow patients to participate and flourish in society. By the same logic, when approached by patients with a request for letters in support of reduced activity, you should consider not only the benefits but also the potential detriments of such disruptions. This may entail recommending temporary restrictions or modifications, as appropriate.

Continue to: Think beyond the patient

Letter writing, particularly when recommending accommodations, can have implications beyond the patient.¹⁶ Such letters may cause unintended societal harm. For example, others may have to assume additional responsibilities; competitive goods (eg, housing) may be rendered to the patient rather than to a person with greater needs; and workplace safety could be compromised due to absence. Consider not only the individual patient but also possible public health and societal effects of letter writing.

Deciding not to write

From an ethical perspective, a physician cannot be compelled to write a letter if such an undertaking violates a stronger moral obligation. An example of this is if writing a letter could cause significant harm to the patient or society, or if writing a letter might compromise a physician’s professionalism.¹⁹ When you elect to not write a letter, the ethical principles of autonomy and truth telling dictate that you must inform your patients of this choice.⁶ You should also provide an explanation to the patient as long as such information would not cause undue psychological or physical harm.^20,21

Schedule time to write letters

Some physicians implement policies that all letters are to be completed during scheduled appointments. Others designate administrative time to complete requested letters. Finally, some physicians flexibly complete such requests between appointments or during other undedicated time slots. Any of these approaches are justifiable, though some urgent requests may require more immediate attention outside of appointments. Some physicians may choose to bill for the letter writing if completed outside an appointment and the patient is treated in private practice. Whatever your policy, inform patients of it at the beginning of care and remind them when appropriate, such as before completing a letter that may be billed.

Manage uncertainty

Always strive for objectivity in letter writing. However, some requests inherently hinge on subjective reports and assessments. For example, a patient may request an excuse letter due to feeling unwell. In the absence of objective findings, what should you do? We advise the following.

Acquire collateral information. Adequate information is essential when making any medical recommendation. The same is true for writing letters. With the patient’s permission, you may need to contact relevant parties to better understand the circumstance or activity about which you are being asked to write a letter. For example, a patient may request leave from work due to injury. If the specific parameters of the work impeded by the injury are unclear to you, refrain from writing the letter and explain the rationale to the patient.

Continue to: Integrate prior knowledge of the patient

Integrate prior knowledge of the patient. No letter writing request exists in a vacuum. If you know the patient, the letter should be contextualized within the patient’s prior behaviors.

Stay within your scope

Given the various dilemmas and challenges, you may want to consider whether some letter writing is out of your professional scope.^14-16 One solution would be to leave such requests to other entities (eg, requiring employers to retain medical personnel with specialized skills in occupational evaluations) and make such recommendations to patients. Regardless, physicians should think carefully about their professional boundaries and scope regarding letter requests and adopt and implement a consistent standard for all patients.

Regarding the letter requested by Ms. M, you should consider whether the appeal is consistent with the patient’s psychiatric illness. You should also consider whether you have sufficient knowledge about the patient’s living environment to support their claim. Such a letter should be written only if you understand both considerations. Regardless of your decision, you should explain your rationale to the patient.

Bottom Line

Patients may ask their psychiatrists to write letters that address aspects of their social well-being. However, psychiatrists must be alert to requests that are outside their scope of practice or ethically or legally fraught. Carefully consider whether writing a letter is appropriate and if not, discuss with the patient the reasons you cannot write such a letter and any recommended alternative avenues to address their request.

Related Resources

• Riese A. Writing letters for transgender patients undergoing medical transition. Current Psychiatry. 2021;20(8):51-52. doi:10.12788/cp.0159
• Joshi KG. Service animals and emotional support animals: should you write that letter? Current Psychiatry. 2021;20(11):16-19,24. doi:10.12788/cp.0183

Approximately 1 in 200 individuals will be diagnosed with schizophrenia in their lifetime.¹ DSM-5 criteria for the diagnosis of schizophrenia require the presence of ≥2 of 5 symptoms: delusions, hallucinations, disordered speech, grossly disorganized (or catatonic) behavior, and negative symptoms such as flat affect or avolition.² Multiple studies have found increased rates of cannabis use among patients with schizophrenia. Because cognitive deficits are the chief predictor of clinical outcomes and quality of life in individuals with schizophrenia, the cognitive effects of cannabis use among these patients are of clinical significance.³ As legislation increasingly allows for the sale, possession, and consumption of cannabis, it is crucial to provide clinicians with evidence-based recommendations for treating patients who regularly use cannabis (approximately 8% of the adult population³). In this article, we analyze several peer-reviewed studies to investigate the impact of cannabis use on the onset and development of schizophrenia.

A look at substance-induced psychosis

Schizophrenia is associated with several structural brain changes, and some of these changes may be influenced by cannabis use (Box⁴). The biochemical etiology of schizophrenia is poorly understood but thought to involve dopamine, glutamate, serotonin, and gamma-aminobutyric acid. Certain positive symptoms, such as hallucinations, are uniquely human and difficult to study in animal models.⁵ Psychoactive substance use, especially cannabis, is frequently comorbid with schizophrenia. Additionally, certain individuals may be more predisposed to substance-induced psychosis than others based on genetic variation and underlying brain structure changes.⁴ Substance-induced psychosis is a psychotic state following the ingestion of a psychoactive substance or drug withdrawal lasting ≥48 hours.⁶ The psychoactive effects of cannabis have been associated with an exacerbation of existing schizophrenia symptoms.⁷ In 1998, Hall⁷ proposed 2 hypotheses to explain the relationship between cannabis and psychosis. The first was that heavy consumption of cannabis triggers a specific type of cannabis psychosis.⁷ The second was that cannabis use exacerbates existing schizophrenia, making the symptoms worse.⁷ Hall⁷ concluded that there was a complicated interaction among an individual’s vulnerability to their stressors, environment, and genetics.

Box

Cannabis, COMT, and homocysteine

Great advances have been made in our ability to examine the association between genetics and metabolism. One example of this is the interaction between the catechol-O-methyltransferase (COMT) gene and the active component of cannabis, delta-9-tetra­hydrocannabinol (THC). COMT codes for an enzyme that degrades cortical dopamine. The Val158Met polymorphism of this gene increases COMT activity, leading to increased dopamine catabolism, and thus decreased levels of extracellular dopamine, which induces an increase in mesolimbic dopaminergic activity, thereby increasing susceptibility to psychosis.³

In a study that genotyped 135 patients with schizophrenia, the Val158Met polymorphism was present in 29.63% of participants.³ Because THC can induce episodes of psychosis, individuals with this polymorphism may be at a higher risk of developing schizophrenia. Compared to Met carrier control participants with similar histories of cannabis consumption, those with the Val158Met polymorphism demonstrated markedly worse performance on tests of verbal fluency and processing speed.³ This is clinically significant because cognitive impairments are a major prognostic factor in schizophrenia, and identifying patients with this polymorphism could help personalize interventions for those who consume cannabis and are at risk of developing schizophrenia.

A study that evaluated 56 patients with first-episode schizophrenia found that having a history of cannabis abuse was associated with significantly higher levels of homocysteine as well as lower levels of high-density lipoprotein and vitamin B12.⁸ Homocysteine is an agonist at the glutamate binding site and a partial antagonist at the glycine co-agonist site in the N-methyl-D-aspartate receptor, which suggests that homocysteine may contribute to hypofunctioning of glutamate transmission; this is implicated in the development of schizophrenia. These increases in homocysteine are also found in siblings of patients with schizophrenia, which indicates a possible association between the methylenetetrahydrofolate (MTHFR) gene and schizophrenia.⁸

The C677T polymorphism in MTHFR may predict the risk of developing metabolic syndrome in patients taking second-generation antipsychotics.⁸ Elevations in homocysteine by as little as 5 μmol/L may increase schizophrenia risk by 70% compared to controls, possibly due to homocysteine initiating neuronal apoptosis, catalyzing dysfunction of the mitochondria, or increasing oxidative stress.⁸ There is a positive correlation between homocysteine levels and severity of negative symptoms (P = .006) and general psychopathology (P = .008) of schizophrenia when analyzed using the Positive and Negative Syndrome Scale.⁸ Negative symptoms such as blunted affect, apathy, anhedonia, and loss of motivation significantly impact the social and economic outcomes of patients diagnosed with schizophrenia.

Research paints a mixed picture

A Danish study analyzed the rates of conversion to schizophrenia or bipolar disorder (BD) among 6,788 individuals who received a diagnosis of substance-induced psychosis from 1994 to 2014.⁶ Ten comparison participants were selected for each case participant, matched on sex and year/month of birth. Participants were followed until the first occurrence of schizophrenia or BD, death, or emigration from Denmark. Substances implicated in the initial psychotic episode included cannabis, alcohol, opioids, sedatives, cocaine, amphetamines, hallucinogens, and combinations of substances.

Continue to: The overall conversion rate...

The overall conversion rate over 20 years was 32.2% (95% CI, 29.7 to 34.9), with 26.0% developing schizophrenia vs 8.4% developing BD.⁶ Of the substances involved, cannabis was the most common, implicated in 41.2% (95% CI, 36.6 to 46.2) of cases.⁶ One-half of male patients converted within 2.0 years and one-half of female patients converted within 4.4 years after a cannabis-induced psychosis.⁶

This study had several limitations. It could not account for any short-term psychotic symptoms experienced by the general population, especially after cannabis use. Such patients might not seek treatment. Thus, the results might not be generalizable to the general population. The study did not evaluate if conversion rates differed based on continued substance use following the psychosis episode, or the amount of each substance taken prior to the episode. Dose-dependence was not well elucidated, and this study only looked at patients from Denmark and did not account for socioeconomic status.⁶

Another Danish study looked at the influences of gender and cannabis use in the early course of the disease in 133 patients with schizophrenia.⁹ These researchers found that male gender was a significant predictor of earlier onset of dysfunction socially and in the workplace, as well as a higher risk of developing negative symptoms. However, compared to gender, cannabis use was a stronger predictor of age at first psychotic episode. For cannabis users, the median age of onset of negative symptoms was 23.7, compared to 38.4 for nonusers (P < .001).⁹

Cannabis use is significantly elevated among individuals with psychosis, with a 12-month prevalence of 29.2% compared to 4.0% among the general population of the United States.¹⁰ In a study that assessed 229 patients with a schizophrenia spectrum disorder during their first hospitalization and 6 months, 2 years, 4 years, and 10 years later, Foti et al¹⁰ found that the lifetime rate of cannabis use was 66.2%. Survival analysis found cannabis use doubled the risk of the onset of psychosis compared to nonusers of the same age (hazard ratio [HR] = 1.97; 95% CI, 1.48 to 2.62, P < .001), even after adjusting for socioeconomic status, age, and gender (HR = 1.34; 95% CI, 1.01 to 1.77, P < .05).¹⁰ Additionally, Foti et al¹⁰ found significant positive correlations between psychotic symptoms and cannabis use in patients with schizophrenia over the course of 10 years. An increase in symptoms was associated with a higher likelihood of cannabis use, and a decrease in symptoms was correlated with a lower likelihood of use (adjusted odds ratio = 1.64; 95% CI, 1.12 to 2.43, P < .0125).¹⁰

Ortiz-Medina et al¹¹ conducted a meta-analysis of 22 studies of 15 cohorts from healthy populations and 12 other cohort follow-up studies that evaluated the onset of psychotic symptoms in individuals who used cannabis. Most studies found associations between cannabis use and the onset of symptoms of schizophrenia, and most determined cannabis was also a major risk factor for other psychotic disorders. Analyses of dose-dependence indicated that repeated cannabis use increased the risk of developing psychotic symptoms. This risk is increased when an individual starts using cannabis before age 15.¹¹ Age seemed to be a stronger predictor of onset and outcome than sex, with no significant differences between men and women. One study in this review found that approximately 8% to 13% cases of schizophrenia may have been solely due to cannabis.¹¹ The most significant limitation to the studies analyzed in this review is that retrospective studies utilize self-reported questionnaires.

Continue to: Other researchers have found...

Other researchers have found it would take a relatively high number of individuals to stop using cannabis to prevent 1 case of schizophrenia. In a study of data from England and Wales, Hickman et al¹² evaluated the best available estimates of the incidence of schizophrenia, rates of heavy and light cannabis use, and risk that cannabis causes schizophrenia to determine the number needed to prevent (NNP) 1 case of schizophrenia. They estimated that it would require approximately 2,800 men age 20 to 24 (90% CI, 2,018 to 4,530) and 4,700 men age 35 to 39 (90% CI, 3,114 to 8,416) who heavily used cannabis to stop their consumption to prevent 1 case of schizophrenia.¹² For women with heavy cannabis use, the mean NNP was 5,470 for women age 25 to 29 (90% CI, 3,640 to 9,839) and 10,870 for women age 35 to 39 (90% CI, 6,786 to 22,732).¹² For light cannabis users, the NNP was 4 to 5 times higher than the NNP for heavy cannabis users. This suggests that clinical interventions aimed at preventing dependence on cannabis would be more effective than interventions aimed at eliminating cannabis use.

Medical cannabis and increased potency

In recent years, the use of medical cannabis, which is used to address adverse effects of chemotherapy as well as neuropathic pain, Parkinson’s disease, and epilepsy, has been increasing.¹³ However, there is a lack of well-conducted randomized clinical trials evaluating medical cannabis’ efficacy and safety. As medical cannabis continues to gain public acceptance and more states permit its legal use, patients and physicians should be fully informed of the known adverse effects, including impaired attention, learning, and motivation.¹³

Several studies have drawn attention to the dose-dependence of many of cannabis’ effects. Since at least the 1960s, the concentration of THC in cannabis has increased substantially, thus increasing its potency. Based on 66,747 samples across 8 studies, 1 meta-analysis estimated that THC concentrations in herbal cannabis increased by 0.29% (P < .001) each year between 1970 and 2017.¹⁴ Similarly, THC concentrations in cannabis resins were found to have increased by 0.57% (P = .017) each year between 1975 and 2017.¹⁴ Cannabis products with high concentrations of THC carry an increased risk of addiction and mental health disorders.¹⁴

Identifying those at highest risk

Despite ongoing research, scientific consensus on the relationship of cannabis to schizophrenia and psychosis has yet to be reached. The disparity between the relatively high prevalence of regular adult use of cannabis (8%⁷)and the low incidence of cannabis-induced psychosis suggests that cannabis use alone is unlikely to lead to episodes of psychosis in individuals who are not predisposed to such episodes. Sarrazin et al¹⁵ evaluated 170 patients with schizophrenia, 31 of whom had cannabis use disorder. They found no significant difference in lifetime symptom dimensions between groups, and proposed that cannabis-associated schizophrenia should not be categorized as a distinct clinical entity of schizophrenia with specific features.¹⁵

Policies that encourage follow-up of patients after episodes of drug-induced psychosis may mitigate the adverse social and economic effects of schizophrenia. Currently, these policies are not widely implemented in health care institutions, possibly because psychotic symptoms may fade after the drug’s effects have dissipated. Despite this, these patients are at high risk of developing schizophrenia and self-harm. New-onset schizophrenia should be promptly identified because delayed diagnosis is associated with worse prognosis.⁶ Additionally, identifying genetic susceptibilities to schizophrenia—such as the Val158Met polymorphisms—in individuals who use cannabis could help clinicians manage or slow the onset or progression of schizophrenia.³ Motivational interviewing strategies should be used to minimize or eliminate cannabis use in individuals with active schizophrenia or psychosis, thus preventing worse outcomes.

Bottom Line

Identifying susceptibilities to schizophrenia may guide interventions in patients who use cannabis. Several large studies have suggested that cannabis use may exacerbate symptoms and worsen the prognosis of schizophrenia. Motivational interviewing strategies aimed at minimizing cannabis use may improve outcomes in patients with schizophrenia.

Related Resources

• Khokhar JY, Dwiel LL, Henricks AM, et al. The link between schizophrenia and substance use disorder: a unifying hypothesis. Schizophr Res. 2018;194:78-85. doi:10.1016/j. schres.2017.04.016
• Otite ES, Solanky A, Doumas S. Adolescents, THC, and the risk of psychosis. Current Psychiatry. 2021;20(12):e1-e2. doi:10.12788/cp.0197

Approximately 1 in 200 individuals will be diagnosed with schizophrenia in their lifetime.¹ DSM-5 criteria for the diagnosis of schizophrenia require the presence of ≥2 of 5 symptoms: delusions, hallucinations, disordered speech, grossly disorganized (or catatonic) behavior, and negative symptoms such as flat affect or avolition.² Multiple studies have found increased rates of cannabis use among patients with schizophrenia. Because cognitive deficits are the chief predictor of clinical outcomes and quality of life in individuals with schizophrenia, the cognitive effects of cannabis use among these patients are of clinical significance.³ As legislation increasingly allows for the sale, possession, and consumption of cannabis, it is crucial to provide clinicians with evidence-based recommendations for treating patients who regularly use cannabis (approximately 8% of the adult population³). In this article, we analyze several peer-reviewed studies to investigate the impact of cannabis use on the onset and development of schizophrenia.

A look at substance-induced psychosis

Schizophrenia is associated with several structural brain changes, and some of these changes may be influenced by cannabis use (Box⁴). The biochemical etiology of schizophrenia is poorly understood but thought to involve dopamine, glutamate, serotonin, and gamma-aminobutyric acid. Certain positive symptoms, such as hallucinations, are uniquely human and difficult to study in animal models.⁵ Psychoactive substance use, especially cannabis, is frequently comorbid with schizophrenia. Additionally, certain individuals may be more predisposed to substance-induced psychosis than others based on genetic variation and underlying brain structure changes.⁴ Substance-induced psychosis is a psychotic state following the ingestion of a psychoactive substance or drug withdrawal lasting ≥48 hours.⁶ The psychoactive effects of cannabis have been associated with an exacerbation of existing schizophrenia symptoms.⁷ In 1998, Hall⁷ proposed 2 hypotheses to explain the relationship between cannabis and psychosis. The first was that heavy consumption of cannabis triggers a specific type of cannabis psychosis.⁷ The second was that cannabis use exacerbates existing schizophrenia, making the symptoms worse.⁷ Hall⁷ concluded that there was a complicated interaction among an individual’s vulnerability to their stressors, environment, and genetics.

Box

Cannabis, COMT, and homocysteine

Great advances have been made in our ability to examine the association between genetics and metabolism. One example of this is the interaction between the catechol-O-methyltransferase (COMT) gene and the active component of cannabis, delta-9-tetra­hydrocannabinol (THC). COMT codes for an enzyme that degrades cortical dopamine. The Val158Met polymorphism of this gene increases COMT activity, leading to increased dopamine catabolism, and thus decreased levels of extracellular dopamine, which induces an increase in mesolimbic dopaminergic activity, thereby increasing susceptibility to psychosis.³

In a study that genotyped 135 patients with schizophrenia, the Val158Met polymorphism was present in 29.63% of participants.³ Because THC can induce episodes of psychosis, individuals with this polymorphism may be at a higher risk of developing schizophrenia. Compared to Met carrier control participants with similar histories of cannabis consumption, those with the Val158Met polymorphism demonstrated markedly worse performance on tests of verbal fluency and processing speed.³ This is clinically significant because cognitive impairments are a major prognostic factor in schizophrenia, and identifying patients with this polymorphism could help personalize interventions for those who consume cannabis and are at risk of developing schizophrenia.

A study that evaluated 56 patients with first-episode schizophrenia found that having a history of cannabis abuse was associated with significantly higher levels of homocysteine as well as lower levels of high-density lipoprotein and vitamin B12.⁸ Homocysteine is an agonist at the glutamate binding site and a partial antagonist at the glycine co-agonist site in the N-methyl-D-aspartate receptor, which suggests that homocysteine may contribute to hypofunctioning of glutamate transmission; this is implicated in the development of schizophrenia. These increases in homocysteine are also found in siblings of patients with schizophrenia, which indicates a possible association between the methylenetetrahydrofolate (MTHFR) gene and schizophrenia.⁸

The C677T polymorphism in MTHFR may predict the risk of developing metabolic syndrome in patients taking second-generation antipsychotics.⁸ Elevations in homocysteine by as little as 5 μmol/L may increase schizophrenia risk by 70% compared to controls, possibly due to homocysteine initiating neuronal apoptosis, catalyzing dysfunction of the mitochondria, or increasing oxidative stress.⁸ There is a positive correlation between homocysteine levels and severity of negative symptoms (P = .006) and general psychopathology (P = .008) of schizophrenia when analyzed using the Positive and Negative Syndrome Scale.⁸ Negative symptoms such as blunted affect, apathy, anhedonia, and loss of motivation significantly impact the social and economic outcomes of patients diagnosed with schizophrenia.

Research paints a mixed picture

A Danish study analyzed the rates of conversion to schizophrenia or bipolar disorder (BD) among 6,788 individuals who received a diagnosis of substance-induced psychosis from 1994 to 2014.⁶ Ten comparison participants were selected for each case participant, matched on sex and year/month of birth. Participants were followed until the first occurrence of schizophrenia or BD, death, or emigration from Denmark. Substances implicated in the initial psychotic episode included cannabis, alcohol, opioids, sedatives, cocaine, amphetamines, hallucinogens, and combinations of substances.

Continue to: The overall conversion rate...

The overall conversion rate over 20 years was 32.2% (95% CI, 29.7 to 34.9), with 26.0% developing schizophrenia vs 8.4% developing BD.⁶ Of the substances involved, cannabis was the most common, implicated in 41.2% (95% CI, 36.6 to 46.2) of cases.⁶ One-half of male patients converted within 2.0 years and one-half of female patients converted within 4.4 years after a cannabis-induced psychosis.⁶

This study had several limitations. It could not account for any short-term psychotic symptoms experienced by the general population, especially after cannabis use. Such patients might not seek treatment. Thus, the results might not be generalizable to the general population. The study did not evaluate if conversion rates differed based on continued substance use following the psychosis episode, or the amount of each substance taken prior to the episode. Dose-dependence was not well elucidated, and this study only looked at patients from Denmark and did not account for socioeconomic status.⁶

Another Danish study looked at the influences of gender and cannabis use in the early course of the disease in 133 patients with schizophrenia.⁹ These researchers found that male gender was a significant predictor of earlier onset of dysfunction socially and in the workplace, as well as a higher risk of developing negative symptoms. However, compared to gender, cannabis use was a stronger predictor of age at first psychotic episode. For cannabis users, the median age of onset of negative symptoms was 23.7, compared to 38.4 for nonusers (P < .001).⁹

Cannabis use is significantly elevated among individuals with psychosis, with a 12-month prevalence of 29.2% compared to 4.0% among the general population of the United States.¹⁰ In a study that assessed 229 patients with a schizophrenia spectrum disorder during their first hospitalization and 6 months, 2 years, 4 years, and 10 years later, Foti et al¹⁰ found that the lifetime rate of cannabis use was 66.2%. Survival analysis found cannabis use doubled the risk of the onset of psychosis compared to nonusers of the same age (hazard ratio [HR] = 1.97; 95% CI, 1.48 to 2.62, P < .001), even after adjusting for socioeconomic status, age, and gender (HR = 1.34; 95% CI, 1.01 to 1.77, P < .05).¹⁰ Additionally, Foti et al¹⁰ found significant positive correlations between psychotic symptoms and cannabis use in patients with schizophrenia over the course of 10 years. An increase in symptoms was associated with a higher likelihood of cannabis use, and a decrease in symptoms was correlated with a lower likelihood of use (adjusted odds ratio = 1.64; 95% CI, 1.12 to 2.43, P < .0125).¹⁰

Ortiz-Medina et al¹¹ conducted a meta-analysis of 22 studies of 15 cohorts from healthy populations and 12 other cohort follow-up studies that evaluated the onset of psychotic symptoms in individuals who used cannabis. Most studies found associations between cannabis use and the onset of symptoms of schizophrenia, and most determined cannabis was also a major risk factor for other psychotic disorders. Analyses of dose-dependence indicated that repeated cannabis use increased the risk of developing psychotic symptoms. This risk is increased when an individual starts using cannabis before age 15.¹¹ Age seemed to be a stronger predictor of onset and outcome than sex, with no significant differences between men and women. One study in this review found that approximately 8% to 13% cases of schizophrenia may have been solely due to cannabis.¹¹ The most significant limitation to the studies analyzed in this review is that retrospective studies utilize self-reported questionnaires.

Continue to: Other researchers have found...

Other researchers have found it would take a relatively high number of individuals to stop using cannabis to prevent 1 case of schizophrenia. In a study of data from England and Wales, Hickman et al¹² evaluated the best available estimates of the incidence of schizophrenia, rates of heavy and light cannabis use, and risk that cannabis causes schizophrenia to determine the number needed to prevent (NNP) 1 case of schizophrenia. They estimated that it would require approximately 2,800 men age 20 to 24 (90% CI, 2,018 to 4,530) and 4,700 men age 35 to 39 (90% CI, 3,114 to 8,416) who heavily used cannabis to stop their consumption to prevent 1 case of schizophrenia.¹² For women with heavy cannabis use, the mean NNP was 5,470 for women age 25 to 29 (90% CI, 3,640 to 9,839) and 10,870 for women age 35 to 39 (90% CI, 6,786 to 22,732).¹² For light cannabis users, the NNP was 4 to 5 times higher than the NNP for heavy cannabis users. This suggests that clinical interventions aimed at preventing dependence on cannabis would be more effective than interventions aimed at eliminating cannabis use.

Medical cannabis and increased potency

In recent years, the use of medical cannabis, which is used to address adverse effects of chemotherapy as well as neuropathic pain, Parkinson’s disease, and epilepsy, has been increasing.¹³ However, there is a lack of well-conducted randomized clinical trials evaluating medical cannabis’ efficacy and safety. As medical cannabis continues to gain public acceptance and more states permit its legal use, patients and physicians should be fully informed of the known adverse effects, including impaired attention, learning, and motivation.¹³

Several studies have drawn attention to the dose-dependence of many of cannabis’ effects. Since at least the 1960s, the concentration of THC in cannabis has increased substantially, thus increasing its potency. Based on 66,747 samples across 8 studies, 1 meta-analysis estimated that THC concentrations in herbal cannabis increased by 0.29% (P < .001) each year between 1970 and 2017.¹⁴ Similarly, THC concentrations in cannabis resins were found to have increased by 0.57% (P = .017) each year between 1975 and 2017.¹⁴ Cannabis products with high concentrations of THC carry an increased risk of addiction and mental health disorders.¹⁴

Identifying those at highest risk

Despite ongoing research, scientific consensus on the relationship of cannabis to schizophrenia and psychosis has yet to be reached. The disparity between the relatively high prevalence of regular adult use of cannabis (8%⁷)and the low incidence of cannabis-induced psychosis suggests that cannabis use alone is unlikely to lead to episodes of psychosis in individuals who are not predisposed to such episodes. Sarrazin et al¹⁵ evaluated 170 patients with schizophrenia, 31 of whom had cannabis use disorder. They found no significant difference in lifetime symptom dimensions between groups, and proposed that cannabis-associated schizophrenia should not be categorized as a distinct clinical entity of schizophrenia with specific features.¹⁵

Policies that encourage follow-up of patients after episodes of drug-induced psychosis may mitigate the adverse social and economic effects of schizophrenia. Currently, these policies are not widely implemented in health care institutions, possibly because psychotic symptoms may fade after the drug’s effects have dissipated. Despite this, these patients are at high risk of developing schizophrenia and self-harm. New-onset schizophrenia should be promptly identified because delayed diagnosis is associated with worse prognosis.⁶ Additionally, identifying genetic susceptibilities to schizophrenia—such as the Val158Met polymorphisms—in individuals who use cannabis could help clinicians manage or slow the onset or progression of schizophrenia.³ Motivational interviewing strategies should be used to minimize or eliminate cannabis use in individuals with active schizophrenia or psychosis, thus preventing worse outcomes.

Bottom Line

Identifying susceptibilities to schizophrenia may guide interventions in patients who use cannabis. Several large studies have suggested that cannabis use may exacerbate symptoms and worsen the prognosis of schizophrenia. Motivational interviewing strategies aimed at minimizing cannabis use may improve outcomes in patients with schizophrenia.

Related Resources

• Khokhar JY, Dwiel LL, Henricks AM, et al. The link between schizophrenia and substance use disorder: a unifying hypothesis. Schizophr Res. 2018;194:78-85. doi:10.1016/j. schres.2017.04.016
• Otite ES, Solanky A, Doumas S. Adolescents, THC, and the risk of psychosis. Current Psychiatry. 2021;20(12):e1-e2. doi:10.12788/cp.0197

Approximately 1 in 200 individuals will be diagnosed with schizophrenia in their lifetime.¹ DSM-5 criteria for the diagnosis of schizophrenia require the presence of ≥2 of 5 symptoms: delusions, hallucinations, disordered speech, grossly disorganized (or catatonic) behavior, and negative symptoms such as flat affect or avolition.² Multiple studies have found increased rates of cannabis use among patients with schizophrenia. Because cognitive deficits are the chief predictor of clinical outcomes and quality of life in individuals with schizophrenia, the cognitive effects of cannabis use among these patients are of clinical significance.³ As legislation increasingly allows for the sale, possession, and consumption of cannabis, it is crucial to provide clinicians with evidence-based recommendations for treating patients who regularly use cannabis (approximately 8% of the adult population³). In this article, we analyze several peer-reviewed studies to investigate the impact of cannabis use on the onset and development of schizophrenia.

A look at substance-induced psychosis

Schizophrenia is associated with several structural brain changes, and some of these changes may be influenced by cannabis use (Box⁴). The biochemical etiology of schizophrenia is poorly understood but thought to involve dopamine, glutamate, serotonin, and gamma-aminobutyric acid. Certain positive symptoms, such as hallucinations, are uniquely human and difficult to study in animal models.⁵ Psychoactive substance use, especially cannabis, is frequently comorbid with schizophrenia. Additionally, certain individuals may be more predisposed to substance-induced psychosis than others based on genetic variation and underlying brain structure changes.⁴ Substance-induced psychosis is a psychotic state following the ingestion of a psychoactive substance or drug withdrawal lasting ≥48 hours.⁶ The psychoactive effects of cannabis have been associated with an exacerbation of existing schizophrenia symptoms.⁷ In 1998, Hall⁷ proposed 2 hypotheses to explain the relationship between cannabis and psychosis. The first was that heavy consumption of cannabis triggers a specific type of cannabis psychosis.⁷ The second was that cannabis use exacerbates existing schizophrenia, making the symptoms worse.⁷ Hall⁷ concluded that there was a complicated interaction among an individual’s vulnerability to their stressors, environment, and genetics.

Box

Cannabis, COMT, and homocysteine

Great advances have been made in our ability to examine the association between genetics and metabolism. One example of this is the interaction between the catechol-O-methyltransferase (COMT) gene and the active component of cannabis, delta-9-tetra­hydrocannabinol (THC). COMT codes for an enzyme that degrades cortical dopamine. The Val158Met polymorphism of this gene increases COMT activity, leading to increased dopamine catabolism, and thus decreased levels of extracellular dopamine, which induces an increase in mesolimbic dopaminergic activity, thereby increasing susceptibility to psychosis.³

In a study that genotyped 135 patients with schizophrenia, the Val158Met polymorphism was present in 29.63% of participants.³ Because THC can induce episodes of psychosis, individuals with this polymorphism may be at a higher risk of developing schizophrenia. Compared to Met carrier control participants with similar histories of cannabis consumption, those with the Val158Met polymorphism demonstrated markedly worse performance on tests of verbal fluency and processing speed.³ This is clinically significant because cognitive impairments are a major prognostic factor in schizophrenia, and identifying patients with this polymorphism could help personalize interventions for those who consume cannabis and are at risk of developing schizophrenia.

A study that evaluated 56 patients with first-episode schizophrenia found that having a history of cannabis abuse was associated with significantly higher levels of homocysteine as well as lower levels of high-density lipoprotein and vitamin B12.⁸ Homocysteine is an agonist at the glutamate binding site and a partial antagonist at the glycine co-agonist site in the N-methyl-D-aspartate receptor, which suggests that homocysteine may contribute to hypofunctioning of glutamate transmission; this is implicated in the development of schizophrenia. These increases in homocysteine are also found in siblings of patients with schizophrenia, which indicates a possible association between the methylenetetrahydrofolate (MTHFR) gene and schizophrenia.⁸

The C677T polymorphism in MTHFR may predict the risk of developing metabolic syndrome in patients taking second-generation antipsychotics.⁸ Elevations in homocysteine by as little as 5 μmol/L may increase schizophrenia risk by 70% compared to controls, possibly due to homocysteine initiating neuronal apoptosis, catalyzing dysfunction of the mitochondria, or increasing oxidative stress.⁸ There is a positive correlation between homocysteine levels and severity of negative symptoms (P = .006) and general psychopathology (P = .008) of schizophrenia when analyzed using the Positive and Negative Syndrome Scale.⁸ Negative symptoms such as blunted affect, apathy, anhedonia, and loss of motivation significantly impact the social and economic outcomes of patients diagnosed with schizophrenia.

Research paints a mixed picture

A Danish study analyzed the rates of conversion to schizophrenia or bipolar disorder (BD) among 6,788 individuals who received a diagnosis of substance-induced psychosis from 1994 to 2014.⁶ Ten comparison participants were selected for each case participant, matched on sex and year/month of birth. Participants were followed until the first occurrence of schizophrenia or BD, death, or emigration from Denmark. Substances implicated in the initial psychotic episode included cannabis, alcohol, opioids, sedatives, cocaine, amphetamines, hallucinogens, and combinations of substances.

Continue to: The overall conversion rate...

The overall conversion rate over 20 years was 32.2% (95% CI, 29.7 to 34.9), with 26.0% developing schizophrenia vs 8.4% developing BD.⁶ Of the substances involved, cannabis was the most common, implicated in 41.2% (95% CI, 36.6 to 46.2) of cases.⁶ One-half of male patients converted within 2.0 years and one-half of female patients converted within 4.4 years after a cannabis-induced psychosis.⁶

This study had several limitations. It could not account for any short-term psychotic symptoms experienced by the general population, especially after cannabis use. Such patients might not seek treatment. Thus, the results might not be generalizable to the general population. The study did not evaluate if conversion rates differed based on continued substance use following the psychosis episode, or the amount of each substance taken prior to the episode. Dose-dependence was not well elucidated, and this study only looked at patients from Denmark and did not account for socioeconomic status.⁶

Another Danish study looked at the influences of gender and cannabis use in the early course of the disease in 133 patients with schizophrenia.⁹ These researchers found that male gender was a significant predictor of earlier onset of dysfunction socially and in the workplace, as well as a higher risk of developing negative symptoms. However, compared to gender, cannabis use was a stronger predictor of age at first psychotic episode. For cannabis users, the median age of onset of negative symptoms was 23.7, compared to 38.4 for nonusers (P < .001).⁹

Cannabis use is significantly elevated among individuals with psychosis, with a 12-month prevalence of 29.2% compared to 4.0% among the general population of the United States.¹⁰ In a study that assessed 229 patients with a schizophrenia spectrum disorder during their first hospitalization and 6 months, 2 years, 4 years, and 10 years later, Foti et al¹⁰ found that the lifetime rate of cannabis use was 66.2%. Survival analysis found cannabis use doubled the risk of the onset of psychosis compared to nonusers of the same age (hazard ratio [HR] = 1.97; 95% CI, 1.48 to 2.62, P < .001), even after adjusting for socioeconomic status, age, and gender (HR = 1.34; 95% CI, 1.01 to 1.77, P < .05).¹⁰ Additionally, Foti et al¹⁰ found significant positive correlations between psychotic symptoms and cannabis use in patients with schizophrenia over the course of 10 years. An increase in symptoms was associated with a higher likelihood of cannabis use, and a decrease in symptoms was correlated with a lower likelihood of use (adjusted odds ratio = 1.64; 95% CI, 1.12 to 2.43, P < .0125).¹⁰

Ortiz-Medina et al¹¹ conducted a meta-analysis of 22 studies of 15 cohorts from healthy populations and 12 other cohort follow-up studies that evaluated the onset of psychotic symptoms in individuals who used cannabis. Most studies found associations between cannabis use and the onset of symptoms of schizophrenia, and most determined cannabis was also a major risk factor for other psychotic disorders. Analyses of dose-dependence indicated that repeated cannabis use increased the risk of developing psychotic symptoms. This risk is increased when an individual starts using cannabis before age 15.¹¹ Age seemed to be a stronger predictor of onset and outcome than sex, with no significant differences between men and women. One study in this review found that approximately 8% to 13% cases of schizophrenia may have been solely due to cannabis.¹¹ The most significant limitation to the studies analyzed in this review is that retrospective studies utilize self-reported questionnaires.

Continue to: Other researchers have found...

Other researchers have found it would take a relatively high number of individuals to stop using cannabis to prevent 1 case of schizophrenia. In a study of data from England and Wales, Hickman et al¹² evaluated the best available estimates of the incidence of schizophrenia, rates of heavy and light cannabis use, and risk that cannabis causes schizophrenia to determine the number needed to prevent (NNP) 1 case of schizophrenia. They estimated that it would require approximately 2,800 men age 20 to 24 (90% CI, 2,018 to 4,530) and 4,700 men age 35 to 39 (90% CI, 3,114 to 8,416) who heavily used cannabis to stop their consumption to prevent 1 case of schizophrenia.¹² For women with heavy cannabis use, the mean NNP was 5,470 for women age 25 to 29 (90% CI, 3,640 to 9,839) and 10,870 for women age 35 to 39 (90% CI, 6,786 to 22,732).¹² For light cannabis users, the NNP was 4 to 5 times higher than the NNP for heavy cannabis users. This suggests that clinical interventions aimed at preventing dependence on cannabis would be more effective than interventions aimed at eliminating cannabis use.

Medical cannabis and increased potency

In recent years, the use of medical cannabis, which is used to address adverse effects of chemotherapy as well as neuropathic pain, Parkinson’s disease, and epilepsy, has been increasing.¹³ However, there is a lack of well-conducted randomized clinical trials evaluating medical cannabis’ efficacy and safety. As medical cannabis continues to gain public acceptance and more states permit its legal use, patients and physicians should be fully informed of the known adverse effects, including impaired attention, learning, and motivation.¹³

Several studies have drawn attention to the dose-dependence of many of cannabis’ effects. Since at least the 1960s, the concentration of THC in cannabis has increased substantially, thus increasing its potency. Based on 66,747 samples across 8 studies, 1 meta-analysis estimated that THC concentrations in herbal cannabis increased by 0.29% (P < .001) each year between 1970 and 2017.¹⁴ Similarly, THC concentrations in cannabis resins were found to have increased by 0.57% (P = .017) each year between 1975 and 2017.¹⁴ Cannabis products with high concentrations of THC carry an increased risk of addiction and mental health disorders.¹⁴

Identifying those at highest risk

Despite ongoing research, scientific consensus on the relationship of cannabis to schizophrenia and psychosis has yet to be reached. The disparity between the relatively high prevalence of regular adult use of cannabis (8%⁷)and the low incidence of cannabis-induced psychosis suggests that cannabis use alone is unlikely to lead to episodes of psychosis in individuals who are not predisposed to such episodes. Sarrazin et al¹⁵ evaluated 170 patients with schizophrenia, 31 of whom had cannabis use disorder. They found no significant difference in lifetime symptom dimensions between groups, and proposed that cannabis-associated schizophrenia should not be categorized as a distinct clinical entity of schizophrenia with specific features.¹⁵

Policies that encourage follow-up of patients after episodes of drug-induced psychosis may mitigate the adverse social and economic effects of schizophrenia. Currently, these policies are not widely implemented in health care institutions, possibly because psychotic symptoms may fade after the drug’s effects have dissipated. Despite this, these patients are at high risk of developing schizophrenia and self-harm. New-onset schizophrenia should be promptly identified because delayed diagnosis is associated with worse prognosis.⁶ Additionally, identifying genetic susceptibilities to schizophrenia—such as the Val158Met polymorphisms—in individuals who use cannabis could help clinicians manage or slow the onset or progression of schizophrenia.³ Motivational interviewing strategies should be used to minimize or eliminate cannabis use in individuals with active schizophrenia or psychosis, thus preventing worse outcomes.

Bottom Line

Identifying susceptibilities to schizophrenia may guide interventions in patients who use cannabis. Several large studies have suggested that cannabis use may exacerbate symptoms and worsen the prognosis of schizophrenia. Motivational interviewing strategies aimed at minimizing cannabis use may improve outcomes in patients with schizophrenia.

Related Resources

• Khokhar JY, Dwiel LL, Henricks AM, et al. The link between schizophrenia and substance use disorder: a unifying hypothesis. Schizophr Res. 2018;194:78-85. doi:10.1016/j. schres.2017.04.016
• Otite ES, Solanky A, Doumas S. Adolescents, THC, and the risk of psychosis. Current Psychiatry. 2021;20(12):e1-e2. doi:10.12788/cp.0197

Mr. B, age 23, is admitted to an inpatient psychiatric unit for depression. During his hospitalization, Mr. B becomes fixated on obtaining specific medications, including controlled substances. He is treated by Dr. M, a psychiatrist early in her training. In a difficult conversation, Dr. M tells Mr. B he will not be prescribed the medications he is requesting and explains why. Mr. B responds by jumping across a table and repeatedly punching Dr. M. Unit staff restrains Mr. B, and Dr. M leaves to seek medical care.

Assaults perpetrated against employees on inpatient psychiatric units are common.¹ Assaults on physicians can occur at any level of training, including during residency.² This is not a new phenomenon: concerns about patients assaulting psychiatrists and other inpatient staff have been reported for decades.^3-5 Most research surrounding this topic has focused on risk factors for violence and prevention.⁶ Research regarding the aftermath of a patient assault and what services an employee requires have primarily centered on nurses.^7,8

Practical guidance for a psychiatrist who has been assaulted and wants to return to work is difficult to find. This article provides strategies to help psychiatrists (and their colleagues) transition back to work after being the victim of a patient assault. While the recommendations we provide can be applied to trainees as well as attending physicians, there are some considerations specific to residents who have been assaulted (Box^9,10).

Box

Having a plan for after an assault

The aftereffects of a patient assault can take a significant toll on the individual who is assaulted. A 2021 article about psychiatric mental health nurses by Dean et al⁸ identified multiple potential repercussions of unaddressed workplace violence, including role confusion, job dissatisfaction, decreased resiliency traits, poor coping methods, increased attrition rate, and increased expenditures related to assault injuries. Providing appropriate services and having a plan for how best to support an assaulted psychiatrist are likely to mitigate these effects. This can be grouped into 4 categories: 1) seeking immediate care, 2) removing the patient from your care, 3) easing back into the environment, and 4) finding long-term support.

1. Seeking immediate care

“Round or be rounded on” is a phrase that encapsulates many physicians’ attitude regarding their own health care and may contribute to their refusal of medical care following acute trauma such as an assault. Feelings of shock, guilt, and shame may also lead to a psychiatrist’s initial hesitation to seek treatment. However, it is important for the victim of an assault to be promptly evaluated and treated.

Elevated adrenaline in the aftermath of a physical engagement may mask the perception of injuries, and there is a risk for exposure to blood-borne pathogens. Regardless of the severity of injuries, seeking medical care establishes documentation of any injuries that can later serve as a record for workers’ compensation claims or if legal action is taken.

In addition to medical needs, immediate psychological support should be considered. Compulsory participation in crisis intervention stress debriefing, particularly when performed by untrained individuals, is not recommended due to questions about its demonstrated efficacy and potential to increase the risk of posttraumatic stress disorder (PTSD) in the long term.^11,12 However, research has established the need for immediate support that does not necessarily involve a discussion of the traumatic event. One option is psychological first aid (PFA), an intervention supported by the World Health Organization. Originally developed for victims of mass crisis events, PFA easily translates to the hospital setting.^12,13 PFA focuses on the immediate, basic needs of the victim to reduce distress and anxiety and encourage adaptive coping. Table 1^12,13 summarizes key components of PFA.

Continue to: PFA can be compared...

PFA can be compared to medical first aid in the field prior to reaching the hospital. In the case of Dr. M, other residents collaborated to transport her to the hospital, keep attendings and program directors apprised of the situation, and bring her snacks and comfort items to the hospital. Dr. M also received support from attending physicians at a neighboring hospital who helped coordinate her care. Essentially, she received a de facto version of PFA. However, given the evidence behind PFA and the unfortunate rate of violence against health care staff, institutions and organizations may offer training in PFA to ensure this level of support for all victims.

Multiple groups may take the lead to support a physician following an injury, including human resources, employee health, or other offices within the institution. The principles of PFA can be used to guide these employees in assisting the victim. Even if such employees are not trained in PFA, they can align with these principles by ensuring access to counseling and medical care, assisting with time off and accommodations, and helping the victim of an assault navigate the legal and administrative processes. Workers’ compensation can be a challenging process, and an institution’s human resources department should be available to assist the assaulted individual in navigating resources both within and outside of what they are able to offer.

2. Removing the patient from the psychiatrist’s care

During her recovery, Dr. M heard from a few peers that what happened was an occupational hazard. On some level, they were correct. While the public does not perceive a career in medicine to be physically dangerous, violence is a rampant problem in health care. Research shows that health care professionals are up to 16 times more likely to experience violence than other occupations; the odds for nurses are even higher.⁸

The frequency and pervasiveness of violence against health care professionals create an environment in which it can become an expected, and even accepted, phenomenon. However, violence cannot and should not be viewed as a normal part of workplace culture. A 2016 study by Moylan et al⁷ found that many nurses believe violence is part of their role, and therefore do not recognize the need to report such incidents or seek the necessary support. In other studies, only 30% of nurses reported violence, and the rate of reporting by physicians was 26%.¹⁴ This underreporting likely represents the role confusion surrounding whether caring for self or caring for the patient takes precedent, as well as normative expectations surrounding violence in the workplace.

It must be made clear to the victim that their safety is a priority and violence will not be tolerated. An institution’s administration can achieve this by immediately removing the patient from the victim’s care. In many cases, discharge of the patient from the clinic or facility may be warranted. A psychiatrist should not be expected to continue as the primary physician for a patient who has assaulted them; transfer to another psychiatrist is necessary if discharge is not an appropriate option. In a scenario in which a psychiatrist must maintain the treating relationship with a patient who assaulted them until the patient can be placed with another clinician (eg, as might occur on a unit with severely limited resources), staff chaperones can be considered when interacting with the patient.

Continue to: An institution's adminstration... 

An institution’s administration should provide support if the psychiatrist chooses to press charges. At the core of our ethos as physicians is “do no harm,” and for some, the prospect of filing charges may be a difficult decision. However, health care professionals do not have an ethical obligation to put themselves in danger of serious bodily harm.¹⁵ While there is no one-size-fits-all answer to the question of whether or not to press charges against a patient who has committed an assault, the Occupational Safety and Health Administration considers the perception that violence is tolerated and victims are unable to report to law enforcement an organizational risk factor for workplace violence.¹⁶

As leaders in the workplace, physicians should set the precedent that violence will not be tolerated by reporting incidents to police and filing charges when appropriate. In the case of Dr. M, she received full support from her institution’s administration in filing charges against Mr. B due to the specific details of the assault.

3. Easing back into the environment

Despite assurances from her superiors that she could take time off, Dr. M wanted to return to work as soon as possible. She considered the balance between her physical injuries and desire to return to work and ultimately returned to work 5 days after the assault. She did well with supportive measures from administration and other staff, including the use of technician escorts on the unit, peer support, and frequent communication with and check-ins from management.

The decision on how quickly to return to work should always lie with the individual who was assaulted. The administration should offer time off without hesitation. Victims of an assault may feel overwhelmed by 2 diverging paths on how to return to a traumatic environment: avoid the location at all costs, or try to “face their fears” and return as quickly as possible. Research from outside medicine indicates that the timing of returning to work after a traumatic injury may not be nearly as important as the method of returning, and who makes this decision.¹⁷ Predictors of return to work after an assault include not only the severity of the trauma and amount of distress symptoms, but also any actual or perceived injustice on the part of the victim.¹⁷ Although this study was not specific to health care employees, it suggests that overall, an employee who does not feel a sense of control over their choice to return to work could perceive that as an injustice on the part of administration, leading to decreased job satisfaction.¹⁷

A study by Lamothe et al¹⁸ that was specific to health care professionals found that despite the importance of self-efficacy for the assault victim, perceived organizational support had an even greater protective effect following patient violence.Additionally, monitoring for signs of distress among victims after an episode of violence could prevent further violence by reducing the risk for subsequent victimization.¹⁸ This highlights the need for leadership of an inpatient unit to be keenly aware of how an assault on a psychiatrist or other health care professional may change the work environment and create a need to help staff navigate the new normal they may face on the unit.

Continue to: Finding long-term support

Longitudinal support is key in the initial transition back to work, as well as in the following weeks and months. Studies assessing the impacts of patient assault on mental health nurses indicate that while most individuals exposed to a traumatic event do not develop PTSD, many reported continued somatic symptoms, and more still reported ongoing psychological effects such as recurring thoughts of the assault, fear, generalized anger, and feeling a loss of control.⁸ Peer support is a common method employed by physicians and nurses alike, but administrative support is also essential.⁸

Regardless which form of psychotherapy, medication treatment, or peer support is utilized, access to the tools the psychiatrist finds most helpful is crucial to making them feel safe and comfortable returning to their role. Table 2 details practical steps administrators and peers can take to facilitate longitudinal support in these situations. In the case of Dr. M, administration was not only supportive in encouraging time off, but also in allowing protected time for therapy when she endorsed distress over the event. The combination of immediate responses and more long-term support greatly helped Dr. M continue her role as a psychiatrist and remain satisfied with her work.

Bottom Line

Being assaulted by a patient can make a psychiatrist reluctant to return to work. Strategies to ease this transition include seeking immediate care, removing the patient from the care of the psychiatrist who was assaulted, easing back into the environment, and finding long-term support.

Related Resources

• Lapic S, Joshi KG. What to do after a patient assaults you. Current Psychiatry. 2017;16(10):53-54.
• Joshi KG. Workplace violence: enhance your safety in outpatient settings. Current Psychiatry. 2021;20(8):37-38. doi:10.12788/cp.0163
• Su D. Harassment of health care workers: a survey. Current Psychiatry. 2021;20(6):48-50. doi:10.12788/cp.0135
• Rozel JS, Wiles C, Amin P. Too close for comfort: when the psychiatrist is stalked. Current Psychiatry. 2022;21(1): 23-28. doi:10.12788/cp.0209

Mr. B, age 23, is admitted to an inpatient psychiatric unit for depression. During his hospitalization, Mr. B becomes fixated on obtaining specific medications, including controlled substances. He is treated by Dr. M, a psychiatrist early in her training. In a difficult conversation, Dr. M tells Mr. B he will not be prescribed the medications he is requesting and explains why. Mr. B responds by jumping across a table and repeatedly punching Dr. M. Unit staff restrains Mr. B, and Dr. M leaves to seek medical care.

Assaults perpetrated against employees on inpatient psychiatric units are common.¹ Assaults on physicians can occur at any level of training, including during residency.² This is not a new phenomenon: concerns about patients assaulting psychiatrists and other inpatient staff have been reported for decades.^3-5 Most research surrounding this topic has focused on risk factors for violence and prevention.⁶ Research regarding the aftermath of a patient assault and what services an employee requires have primarily centered on nurses.^7,8

Practical guidance for a psychiatrist who has been assaulted and wants to return to work is difficult to find. This article provides strategies to help psychiatrists (and their colleagues) transition back to work after being the victim of a patient assault. While the recommendations we provide can be applied to trainees as well as attending physicians, there are some considerations specific to residents who have been assaulted (Box^9,10).

Box

Having a plan for after an assault

The aftereffects of a patient assault can take a significant toll on the individual who is assaulted. A 2021 article about psychiatric mental health nurses by Dean et al⁸ identified multiple potential repercussions of unaddressed workplace violence, including role confusion, job dissatisfaction, decreased resiliency traits, poor coping methods, increased attrition rate, and increased expenditures related to assault injuries. Providing appropriate services and having a plan for how best to support an assaulted psychiatrist are likely to mitigate these effects. This can be grouped into 4 categories: 1) seeking immediate care, 2) removing the patient from your care, 3) easing back into the environment, and 4) finding long-term support.

1. Seeking immediate care

“Round or be rounded on” is a phrase that encapsulates many physicians’ attitude regarding their own health care and may contribute to their refusal of medical care following acute trauma such as an assault. Feelings of shock, guilt, and shame may also lead to a psychiatrist’s initial hesitation to seek treatment. However, it is important for the victim of an assault to be promptly evaluated and treated.

Elevated adrenaline in the aftermath of a physical engagement may mask the perception of injuries, and there is a risk for exposure to blood-borne pathogens. Regardless of the severity of injuries, seeking medical care establishes documentation of any injuries that can later serve as a record for workers’ compensation claims or if legal action is taken.

In addition to medical needs, immediate psychological support should be considered. Compulsory participation in crisis intervention stress debriefing, particularly when performed by untrained individuals, is not recommended due to questions about its demonstrated efficacy and potential to increase the risk of posttraumatic stress disorder (PTSD) in the long term.^11,12 However, research has established the need for immediate support that does not necessarily involve a discussion of the traumatic event. One option is psychological first aid (PFA), an intervention supported by the World Health Organization. Originally developed for victims of mass crisis events, PFA easily translates to the hospital setting.^12,13 PFA focuses on the immediate, basic needs of the victim to reduce distress and anxiety and encourage adaptive coping. Table 1^12,13 summarizes key components of PFA.

Continue to: PFA can be compared...

PFA can be compared to medical first aid in the field prior to reaching the hospital. In the case of Dr. M, other residents collaborated to transport her to the hospital, keep attendings and program directors apprised of the situation, and bring her snacks and comfort items to the hospital. Dr. M also received support from attending physicians at a neighboring hospital who helped coordinate her care. Essentially, she received a de facto version of PFA. However, given the evidence behind PFA and the unfortunate rate of violence against health care staff, institutions and organizations may offer training in PFA to ensure this level of support for all victims.

Multiple groups may take the lead to support a physician following an injury, including human resources, employee health, or other offices within the institution. The principles of PFA can be used to guide these employees in assisting the victim. Even if such employees are not trained in PFA, they can align with these principles by ensuring access to counseling and medical care, assisting with time off and accommodations, and helping the victim of an assault navigate the legal and administrative processes. Workers’ compensation can be a challenging process, and an institution’s human resources department should be available to assist the assaulted individual in navigating resources both within and outside of what they are able to offer.

2. Removing the patient from the psychiatrist’s care

During her recovery, Dr. M heard from a few peers that what happened was an occupational hazard. On some level, they were correct. While the public does not perceive a career in medicine to be physically dangerous, violence is a rampant problem in health care. Research shows that health care professionals are up to 16 times more likely to experience violence than other occupations; the odds for nurses are even higher.⁸

The frequency and pervasiveness of violence against health care professionals create an environment in which it can become an expected, and even accepted, phenomenon. However, violence cannot and should not be viewed as a normal part of workplace culture. A 2016 study by Moylan et al⁷ found that many nurses believe violence is part of their role, and therefore do not recognize the need to report such incidents or seek the necessary support. In other studies, only 30% of nurses reported violence, and the rate of reporting by physicians was 26%.¹⁴ This underreporting likely represents the role confusion surrounding whether caring for self or caring for the patient takes precedent, as well as normative expectations surrounding violence in the workplace.

It must be made clear to the victim that their safety is a priority and violence will not be tolerated. An institution’s administration can achieve this by immediately removing the patient from the victim’s care. In many cases, discharge of the patient from the clinic or facility may be warranted. A psychiatrist should not be expected to continue as the primary physician for a patient who has assaulted them; transfer to another psychiatrist is necessary if discharge is not an appropriate option. In a scenario in which a psychiatrist must maintain the treating relationship with a patient who assaulted them until the patient can be placed with another clinician (eg, as might occur on a unit with severely limited resources), staff chaperones can be considered when interacting with the patient.

Continue to: An institution's adminstration... 

An institution’s administration should provide support if the psychiatrist chooses to press charges. At the core of our ethos as physicians is “do no harm,” and for some, the prospect of filing charges may be a difficult decision. However, health care professionals do not have an ethical obligation to put themselves in danger of serious bodily harm.¹⁵ While there is no one-size-fits-all answer to the question of whether or not to press charges against a patient who has committed an assault, the Occupational Safety and Health Administration considers the perception that violence is tolerated and victims are unable to report to law enforcement an organizational risk factor for workplace violence.¹⁶

As leaders in the workplace, physicians should set the precedent that violence will not be tolerated by reporting incidents to police and filing charges when appropriate. In the case of Dr. M, she received full support from her institution’s administration in filing charges against Mr. B due to the specific details of the assault.

3. Easing back into the environment

Despite assurances from her superiors that she could take time off, Dr. M wanted to return to work as soon as possible. She considered the balance between her physical injuries and desire to return to work and ultimately returned to work 5 days after the assault. She did well with supportive measures from administration and other staff, including the use of technician escorts on the unit, peer support, and frequent communication with and check-ins from management.

The decision on how quickly to return to work should always lie with the individual who was assaulted. The administration should offer time off without hesitation. Victims of an assault may feel overwhelmed by 2 diverging paths on how to return to a traumatic environment: avoid the location at all costs, or try to “face their fears” and return as quickly as possible. Research from outside medicine indicates that the timing of returning to work after a traumatic injury may not be nearly as important as the method of returning, and who makes this decision.¹⁷ Predictors of return to work after an assault include not only the severity of the trauma and amount of distress symptoms, but also any actual or perceived injustice on the part of the victim.¹⁷ Although this study was not specific to health care employees, it suggests that overall, an employee who does not feel a sense of control over their choice to return to work could perceive that as an injustice on the part of administration, leading to decreased job satisfaction.¹⁷

A study by Lamothe et al¹⁸ that was specific to health care professionals found that despite the importance of self-efficacy for the assault victim, perceived organizational support had an even greater protective effect following patient violence.Additionally, monitoring for signs of distress among victims after an episode of violence could prevent further violence by reducing the risk for subsequent victimization.¹⁸ This highlights the need for leadership of an inpatient unit to be keenly aware of how an assault on a psychiatrist or other health care professional may change the work environment and create a need to help staff navigate the new normal they may face on the unit.

Continue to: Finding long-term support

Longitudinal support is key in the initial transition back to work, as well as in the following weeks and months. Studies assessing the impacts of patient assault on mental health nurses indicate that while most individuals exposed to a traumatic event do not develop PTSD, many reported continued somatic symptoms, and more still reported ongoing psychological effects such as recurring thoughts of the assault, fear, generalized anger, and feeling a loss of control.⁸ Peer support is a common method employed by physicians and nurses alike, but administrative support is also essential.⁸

Regardless which form of psychotherapy, medication treatment, or peer support is utilized, access to the tools the psychiatrist finds most helpful is crucial to making them feel safe and comfortable returning to their role. Table 2 details practical steps administrators and peers can take to facilitate longitudinal support in these situations. In the case of Dr. M, administration was not only supportive in encouraging time off, but also in allowing protected time for therapy when she endorsed distress over the event. The combination of immediate responses and more long-term support greatly helped Dr. M continue her role as a psychiatrist and remain satisfied with her work.

Bottom Line

Being assaulted by a patient can make a psychiatrist reluctant to return to work. Strategies to ease this transition include seeking immediate care, removing the patient from the care of the psychiatrist who was assaulted, easing back into the environment, and finding long-term support.

Related Resources

• Lapic S, Joshi KG. What to do after a patient assaults you. Current Psychiatry. 2017;16(10):53-54.
• Joshi KG. Workplace violence: enhance your safety in outpatient settings. Current Psychiatry. 2021;20(8):37-38. doi:10.12788/cp.0163
• Su D. Harassment of health care workers: a survey. Current Psychiatry. 2021;20(6):48-50. doi:10.12788/cp.0135
• Rozel JS, Wiles C, Amin P. Too close for comfort: when the psychiatrist is stalked. Current Psychiatry. 2022;21(1): 23-28. doi:10.12788/cp.0209

Mr. B, age 23, is admitted to an inpatient psychiatric unit for depression. During his hospitalization, Mr. B becomes fixated on obtaining specific medications, including controlled substances. He is treated by Dr. M, a psychiatrist early in her training. In a difficult conversation, Dr. M tells Mr. B he will not be prescribed the medications he is requesting and explains why. Mr. B responds by jumping across a table and repeatedly punching Dr. M. Unit staff restrains Mr. B, and Dr. M leaves to seek medical care.

Assaults perpetrated against employees on inpatient psychiatric units are common.¹ Assaults on physicians can occur at any level of training, including during residency.² This is not a new phenomenon: concerns about patients assaulting psychiatrists and other inpatient staff have been reported for decades.^3-5 Most research surrounding this topic has focused on risk factors for violence and prevention.⁶ Research regarding the aftermath of a patient assault and what services an employee requires have primarily centered on nurses.^7,8

Practical guidance for a psychiatrist who has been assaulted and wants to return to work is difficult to find. This article provides strategies to help psychiatrists (and their colleagues) transition back to work after being the victim of a patient assault. While the recommendations we provide can be applied to trainees as well as attending physicians, there are some considerations specific to residents who have been assaulted (Box^9,10).

Box

Having a plan for after an assault

The aftereffects of a patient assault can take a significant toll on the individual who is assaulted. A 2021 article about psychiatric mental health nurses by Dean et al⁸ identified multiple potential repercussions of unaddressed workplace violence, including role confusion, job dissatisfaction, decreased resiliency traits, poor coping methods, increased attrition rate, and increased expenditures related to assault injuries. Providing appropriate services and having a plan for how best to support an assaulted psychiatrist are likely to mitigate these effects. This can be grouped into 4 categories: 1) seeking immediate care, 2) removing the patient from your care, 3) easing back into the environment, and 4) finding long-term support.

1. Seeking immediate care

“Round or be rounded on” is a phrase that encapsulates many physicians’ attitude regarding their own health care and may contribute to their refusal of medical care following acute trauma such as an assault. Feelings of shock, guilt, and shame may also lead to a psychiatrist’s initial hesitation to seek treatment. However, it is important for the victim of an assault to be promptly evaluated and treated.

Elevated adrenaline in the aftermath of a physical engagement may mask the perception of injuries, and there is a risk for exposure to blood-borne pathogens. Regardless of the severity of injuries, seeking medical care establishes documentation of any injuries that can later serve as a record for workers’ compensation claims or if legal action is taken.

In addition to medical needs, immediate psychological support should be considered. Compulsory participation in crisis intervention stress debriefing, particularly when performed by untrained individuals, is not recommended due to questions about its demonstrated efficacy and potential to increase the risk of posttraumatic stress disorder (PTSD) in the long term.^11,12 However, research has established the need for immediate support that does not necessarily involve a discussion of the traumatic event. One option is psychological first aid (PFA), an intervention supported by the World Health Organization. Originally developed for victims of mass crisis events, PFA easily translates to the hospital setting.^12,13 PFA focuses on the immediate, basic needs of the victim to reduce distress and anxiety and encourage adaptive coping. Table 1^12,13 summarizes key components of PFA.

Continue to: PFA can be compared...

PFA can be compared to medical first aid in the field prior to reaching the hospital. In the case of Dr. M, other residents collaborated to transport her to the hospital, keep attendings and program directors apprised of the situation, and bring her snacks and comfort items to the hospital. Dr. M also received support from attending physicians at a neighboring hospital who helped coordinate her care. Essentially, she received a de facto version of PFA. However, given the evidence behind PFA and the unfortunate rate of violence against health care staff, institutions and organizations may offer training in PFA to ensure this level of support for all victims.

Multiple groups may take the lead to support a physician following an injury, including human resources, employee health, or other offices within the institution. The principles of PFA can be used to guide these employees in assisting the victim. Even if such employees are not trained in PFA, they can align with these principles by ensuring access to counseling and medical care, assisting with time off and accommodations, and helping the victim of an assault navigate the legal and administrative processes. Workers’ compensation can be a challenging process, and an institution’s human resources department should be available to assist the assaulted individual in navigating resources both within and outside of what they are able to offer.

2. Removing the patient from the psychiatrist’s care

During her recovery, Dr. M heard from a few peers that what happened was an occupational hazard. On some level, they were correct. While the public does not perceive a career in medicine to be physically dangerous, violence is a rampant problem in health care. Research shows that health care professionals are up to 16 times more likely to experience violence than other occupations; the odds for nurses are even higher.⁸

The frequency and pervasiveness of violence against health care professionals create an environment in which it can become an expected, and even accepted, phenomenon. However, violence cannot and should not be viewed as a normal part of workplace culture. A 2016 study by Moylan et al⁷ found that many nurses believe violence is part of their role, and therefore do not recognize the need to report such incidents or seek the necessary support. In other studies, only 30% of nurses reported violence, and the rate of reporting by physicians was 26%.¹⁴ This underreporting likely represents the role confusion surrounding whether caring for self or caring for the patient takes precedent, as well as normative expectations surrounding violence in the workplace.

It must be made clear to the victim that their safety is a priority and violence will not be tolerated. An institution’s administration can achieve this by immediately removing the patient from the victim’s care. In many cases, discharge of the patient from the clinic or facility may be warranted. A psychiatrist should not be expected to continue as the primary physician for a patient who has assaulted them; transfer to another psychiatrist is necessary if discharge is not an appropriate option. In a scenario in which a psychiatrist must maintain the treating relationship with a patient who assaulted them until the patient can be placed with another clinician (eg, as might occur on a unit with severely limited resources), staff chaperones can be considered when interacting with the patient.

Continue to: An institution's adminstration... 

An institution’s administration should provide support if the psychiatrist chooses to press charges. At the core of our ethos as physicians is “do no harm,” and for some, the prospect of filing charges may be a difficult decision. However, health care professionals do not have an ethical obligation to put themselves in danger of serious bodily harm.¹⁵ While there is no one-size-fits-all answer to the question of whether or not to press charges against a patient who has committed an assault, the Occupational Safety and Health Administration considers the perception that violence is tolerated and victims are unable to report to law enforcement an organizational risk factor for workplace violence.¹⁶

As leaders in the workplace, physicians should set the precedent that violence will not be tolerated by reporting incidents to police and filing charges when appropriate. In the case of Dr. M, she received full support from her institution’s administration in filing charges against Mr. B due to the specific details of the assault.

3. Easing back into the environment

Despite assurances from her superiors that she could take time off, Dr. M wanted to return to work as soon as possible. She considered the balance between her physical injuries and desire to return to work and ultimately returned to work 5 days after the assault. She did well with supportive measures from administration and other staff, including the use of technician escorts on the unit, peer support, and frequent communication with and check-ins from management.

The decision on how quickly to return to work should always lie with the individual who was assaulted. The administration should offer time off without hesitation. Victims of an assault may feel overwhelmed by 2 diverging paths on how to return to a traumatic environment: avoid the location at all costs, or try to “face their fears” and return as quickly as possible. Research from outside medicine indicates that the timing of returning to work after a traumatic injury may not be nearly as important as the method of returning, and who makes this decision.¹⁷ Predictors of return to work after an assault include not only the severity of the trauma and amount of distress symptoms, but also any actual or perceived injustice on the part of the victim.¹⁷ Although this study was not specific to health care employees, it suggests that overall, an employee who does not feel a sense of control over their choice to return to work could perceive that as an injustice on the part of administration, leading to decreased job satisfaction.¹⁷

A study by Lamothe et al¹⁸ that was specific to health care professionals found that despite the importance of self-efficacy for the assault victim, perceived organizational support had an even greater protective effect following patient violence.Additionally, monitoring for signs of distress among victims after an episode of violence could prevent further violence by reducing the risk for subsequent victimization.¹⁸ This highlights the need for leadership of an inpatient unit to be keenly aware of how an assault on a psychiatrist or other health care professional may change the work environment and create a need to help staff navigate the new normal they may face on the unit.

Continue to: Finding long-term support

Longitudinal support is key in the initial transition back to work, as well as in the following weeks and months. Studies assessing the impacts of patient assault on mental health nurses indicate that while most individuals exposed to a traumatic event do not develop PTSD, many reported continued somatic symptoms, and more still reported ongoing psychological effects such as recurring thoughts of the assault, fear, generalized anger, and feeling a loss of control.⁸ Peer support is a common method employed by physicians and nurses alike, but administrative support is also essential.⁸

Regardless which form of psychotherapy, medication treatment, or peer support is utilized, access to the tools the psychiatrist finds most helpful is crucial to making them feel safe and comfortable returning to their role. Table 2 details practical steps administrators and peers can take to facilitate longitudinal support in these situations. In the case of Dr. M, administration was not only supportive in encouraging time off, but also in allowing protected time for therapy when she endorsed distress over the event. The combination of immediate responses and more long-term support greatly helped Dr. M continue her role as a psychiatrist and remain satisfied with her work.

Bottom Line

Being assaulted by a patient can make a psychiatrist reluctant to return to work. Strategies to ease this transition include seeking immediate care, removing the patient from the care of the psychiatrist who was assaulted, easing back into the environment, and finding long-term support.

Related Resources

• Lapic S, Joshi KG. What to do after a patient assaults you. Current Psychiatry. 2017;16(10):53-54.
• Joshi KG. Workplace violence: enhance your safety in outpatient settings. Current Psychiatry. 2021;20(8):37-38. doi:10.12788/cp.0163
• Su D. Harassment of health care workers: a survey. Current Psychiatry. 2021;20(6):48-50. doi:10.12788/cp.0135
• Rozel JS, Wiles C, Amin P. Too close for comfort: when the psychiatrist is stalked. Current Psychiatry. 2022;21(1): 23-28. doi:10.12788/cp.0209

Attention-deficit/hyperactivity disorder (ADHD) is common, with an estimated worldwide prevalence of 5.29% among children and adolescents and 2.5% among adults.¹ DSM-5-TR classifies ADHD as a neurodevelop­mental disorder, “a group of conditions with onset in the developmental period [that] typically manifest early in development, often before the child enters school.”² Because of the expectation that ADHD symptoms emerge early in development, the diagnostic criteria specify that symptoms must have been present prior to age 12 to qualify as ADHD. However, recent years have shown a significant increase in the number of patients being diagnosed with ADHD for the first time in adulthood. One study found that the diagnosis of ADHD among adults in the United States doubled between 2007 and 2016.³

First-line treatment for ADHD is the stimulants methylphenidate and amphetamine/dextroamphetamine. In the United States, these medications are classified as Schedule II controlled substances, indicating a high risk for abuse. However, just as ADHD diagnoses among adults have increased, so have prescriptions for stimulants. For example, Olfson et al⁴ found that stimulant prescriptions among young adults increased by a factor of 10 between 1994 and 2009.

The increased prevalence of adult patients diagnosed with ADHD and taking stimulants frequently places clinicians in a position to consider the validity of existing diagnoses and evaluate new patients with ADHD-related concerns. In this article, we review some of the challenges associated with diagnosing ADHD in adults, discuss the risks of stimulant treatment, and present a practical approach to the diagnosis and treatment of ADHD in adults.

Challenges in diagnosis

DSM-5-TR diagnostic criteria for ADHD are summarized in Table 1. Establishing a diagnosis of adult ADHD can be challenging. As with many psychiatric conditions, symptoms of ADHD are highly subjective. Retrospectively diagnosing a developmental condition in adults is often biased by the patient’s current functioning.⁵ ADHD has a high heritability and adults may inquire about the diagnosis if their children are diagnosed with ADHD.⁶ Some experts have cautioned that clinicians must be careful in diagnosing ADHD in adults.⁷ Just as there are risks associated with underdiagnosing ADHD, there are risks associated with overdiagnosis. Overdiagnosis may medicalize normal variants in the population and lead to unnecessary treatment and a misappropriation of limited medical resources.⁸ Many false positive cases of late-onset ADHD may be attributable to nonimpairing cognitive fluctuations.⁹

Poor diagnostic practices can impede accuracy in establishing the presence or absence of ADHD. Unfortunately, methods of diagnosing adult ADHD have been shown to vary widely in terms of information sources, diagnostic instruments used, symptom threshold, and whether functional impairment is a requirement for diagnosis.¹⁰ A common practice in diagnosing adult ADHD involves asking patients to complete self-report questionnaires that list symptoms of ADHD, such as the Adult ADHD Self-Report Scale developed by the World Health Organization.¹¹ However, self-reports of ADHD in adults are less reliable than informant reports, and some young adults without ADHD overreport symptoms.^12,13 Symptom checklists are particularly susceptible to faking, which lessens their diagnostic value.¹⁴

The possibility of malingered symptoms of ADHD further increases the diagnostic difficulty. College students may be particularly susceptible to overreporting ADHD symptoms in order to obtain academic accommodations or stimulants in the hopes of improving school performance.¹⁵ One study found that 25% to 48% of college students self-referred for ADHD evaluations exaggerated their symptoms.¹⁶ In another study, 31% of adults failed the Word Memory Test, which suggests noncredible performance in their ADHD evaluation.¹⁷ College students can successfully feign ADHD symptoms in both self-reported symptoms and computer-based tests of attention.¹⁸ Harrison et al¹⁹ summarized many of these concerns in their 2007 study of ADHD malingering, noting the “almost perfect ability of the Faking group to choose items … that correspond to the DSM-IV symptoms, and to report these at levels even higher than persons with diagnosed ADHD.” They suggested “Clinicians should be suspicious of students or young adults presenting for a first-time diagnosis who rate themselves as being significantly symptomatic, yet have managed to achieve well in school and other life activities.”¹⁹

Another challenge in correctly diagnosing adult ADHD is identifying other conditions that may impair attention.²⁰ Psychiatric conditions that may impair concentration include anxiety disorders, chronic stress, posttraumatic stress disorder, recent trauma, major depressive disorder (MDD), and bipolar disorder (BD). Undiagnosed learning disorders may present like ADHD. Focus can be negatively affected by sleep disorders such as sleep apnea, restless leg syndrome, or delayed sleep phase-onset disorder. Marijuana, cocaine, 3,4-methylenedioxy-methamphetamine (MDMA; “ecstasy”), caffeine, or prescription medications such as anticholinergics can also impair attention. Medical conditions that can present with attentional or executive functioning deficits include seizures, Lyme disease, HIV, encephalopathy, hypothyroidism, and “chemo brain.”²¹ Environmental factors such as age-related cognitive decline, sleep deprivation, inflammation, obesity, air pollution, chemical exposure, and excessive use of digital media may also produce symptoms similar to ADHD. Two studies of adult-onset ADHD concluded that 93% to 95% of cases were better explained by other conditions such as sleep disorders, substance use disorders, or another psychiatric disorder.²²

Continue to: Risks associated with treatment

Risks associated with treatment

With or without an accurate ADHD diagnosis, prescribing stimulants presents certain risks (Table 2^23-40). One of the more well-known risks of stimulants is addiction or misuse.²³ An estimated 5 million American adults misused prescription stimulants in 2016.²⁴ Despite stimulants’ status as controlled substances, long-term concurrent use of stimulants with opioids is common among adults with ADHD.²⁵ College students are particularly susceptible to misusing or diverting stimulants, often to improve their academic performance.²⁶ At 1 university, 22% of students had misused stimulants in the past year.²⁷ Prescribing short-acting stimulants (rather than extended-release formulations) increases the likelihood of misuse.²⁸ Patients prescribed stimulants begin to receive requests to divert their medications to others as early as elementary school, and by college more than one-third of those taking stimulants have been asked to give, sell, or trade their medications.²⁹ Diversion of stimulants by students with ADHD is prevalent, with 62% of patients engaging in diversion during their lifetime.¹⁵ Diverted stimulants can come from family members, black market sources, or deceived clinicians.³⁰ Although students’ stimulant misuse/diversion often is academically motivated, nonmedical use of psychostimulants does not appear to have a statistically significant effect on improving grade point average.³¹ Despite a negligible impact on grades, most students who take stimulants identify their effect as strongly positive, producing a situation in which misusers of stimulants have little motivation to stop.³² While some patients might ask for a stimulant prescription with the rationale that liking the effects proves they have ADHD, this is inappropriate because most individuals like the effects of stimulant medications.³³

The use of stimulants increases the risk for several adverse psychiatric outcomes. Stimulants increase the risk of anxiety, so exercise caution when prescribing to patients with a comorbid anxiety disorder.³⁴ Stimulants can also worsen irritability and insomnia, 2 issues common among patients with ADHD.³² Use of stimulant medications can trigger manic episodes. Viktorin et al³⁵ found a >6-fold increase in manic episodes among patients with BD receiving methylphenidate monotherapy compared to those receiving a combination of methyl­phenidate and a mood stabilizer.³⁵ The use of methylphenidate and amphetamine can lead to new-onset psychosis (or exacerbation of pre-existing psychotic illness); amphetamine use is associated with a higher risk of psychosis than methylphenidate.³⁶

General medical adverse effects are also possible with stimulant use. Stimulants’ adverse effect profiles include appetite suppression, dry mouth, and nausea. Long-term use poses a risk for stunting growth in children.¹ Using stimulants during pregnancy is associated with higher risk for neonatal morbidity, including preterm birth, CNS-related disorders, and seizures.³⁷ Stimulants can raise blood pressure and increase heart rate. Serious cardiovascular events associated with stimulant use include ventricular arrhythmias, strokes, and transient ischemic attacks.³⁸

Nonstimulant ADHD treatments are less risky than stimulants but still require monitoring for common adverse effects. Atomoxetine has been associated with sedation, growth retardation (in children), and in severe cases, liver injury or suicidal ideation.³⁹ Bupropion (commonly used off-label for ADHD) can lower the seizure threshold and cause irritability, anorexia, and insomnia.³⁹ Viloxazine, a newer agent, can cause hypertension, increased heart rate, nausea, drowsiness, headache, and insomnia.⁴⁰

Sensible diagnosing

Given the challenges in accurately diagnosing ADHD in adults, we present a sensible approach to making the diagnosis (Table 3). The first step is to rule out other conditions that might better explain the patient’s symptoms. A thorough clinical interview (including a psychiatric review of symptoms) is the cornerstone of an initial diagnostic assessment. The use of validated screening questionnaires such as the Patient Health Questionnaire-9 and General Anxiety Disorder-7 may also provide information regarding psychiatric conditions that require additional evaluation.

Continue to: Some of the most common conditions...

Some of the most common conditions we see mistaken for ADHD are MDD, generalized anxiety disorder (GAD), and BD. In DSM-5-TR, 1 of the diagnostic criteria for MDD is “diminished ability to think or concentrate, or indecisiveness, nearly every day (either by subjective account or as observed by others).”⁴¹ Similarly, criteria for GAD include “difficulty concentrating.”⁴² DSM-5-TR also includes distractibility as one of the criteria for mania/hypomania. Table 4^20-22,41,42 lists other psychiatric, substance-related, medical, and environmental conditions that can produce ADHD-like symptoms. Referring to some medical and environmental explanations for inattention, Aiken²² pointed out, “Patients who suffer from these problems might ask their doctor for a stimulant, but none of those syndromes require a psychopharmacologic approach.” ADHD can be comorbid with other psychiatric conditions, so the presence of another psychiatric illness does not automatically rule out ADHD. If alternative psychiatric diagnoses have been identified, these can be discussed with the patient and treatment offered that targets the specified condition.

Once alternative explanations have been ruled out, focus on the patient’s developmental history. DSM-5-TR conceptualizes ADHD as a neurodevelopmental disorder, meaning it is expected to emerge early in life. Whereas previous editions of DSM specified that ADHD symptoms must be present before age 7, DSM-5 modified this age threshold to before age 12.¹ This necessitates taking a careful life history in order to understand the presence or absence of symptoms at earlier developmental stages.⁵ ADHD should be verified by symptoms apparent in childhood and present across the lifespan.¹⁵

While this retrospective history is necessary, histories that rely on self-report alone are often unreliable. Collateral sources of information are generally more reliable when assessing for ADHD symptoms.¹³ Third-party sources can help confirm that any impairment is best attributed to ADHD rather than to another condition.¹⁵ Unfortunately, the difficulty of obtaining collateral information means it is often neglected, even in the literature.¹⁰ A parent is the ideal informant for gathering collateral information regarding a patient’s functioning in childhood.⁵ Suggested best practices also include obtaining collateral information from interviews with significant others, behavioral questionnaires completed by parents (for current and childhood symptoms), review of school records, and consideration of intellectual and achievement testing.⁴³ If psychological testing is pursued, include validity testing to detect feigned symptoms.^18,44

When evaluating for ADHD, assess not only for the presence of symptoms, but also if these symptoms produce significant functional impairment.^13,15 Impairments in daily functioning can include impaired school participation, social participation, quality of relationships, family conflict, family activities, family functioning, and emotional functioning.⁴⁵ Some symptoms may affect functioning in an adult’s life differently than they did during childhood, from missed work appointments to being late picking up kids from school. Research has shown that the correlation between the number of symptoms and functional impairment is weak, which means someone could experience all of the symptoms of ADHD without experiencing functional impairment.⁴⁵ To make an accurate diagnosis, it is therefore important to clearly establish both the number of symptoms the patient is experiencing and whether these symptoms are clearly linked to functional impairments.¹⁰

Sensible treatment

Once a diagnosis of ADHD has been clearly established, clinicians need to consider how best to treat the condition (Table 5). Stimulants are generally considered first-line treatment for ADHD. In randomized clinical trials, they showed significant efficacy; for example, one study of 146 adults with ADHD found a 76% improvement with methylphenidate compared to 19% for the placebo group.⁴⁶ Before starting a stimulant, certain comorbidities should be ruled out. If a patient has glaucoma or pheochromocytoma, they may first need treatment from or clearance by other specialists. Stimulants should likely be held in patients with hypertension, angina, or cardiovascular defects until receiving medical clearance. The risks of stimulants need to be discussed with female patients of childbearing age, weighing the benefits of treatment against the risks of medication use should the patient get pregnant. Patients with comorbid psychosis or uncontrolled bipolar illness should not receive stimulants due to the risk of exacerbation. Patients with active substance use disorders (SUDs) are generally not good candidates for stimulants because of the risk of misusing or diverting stimulants and the possibility that substance abuse may be causing their inattentive symptoms. Patients whose SUDs are in remission may cautiously be considered as candidates for stimulants. If patients misuse their prescribed stimulants, they should be switched to a nonstimulant medication such as atomoxetine, bupropion, guanfacine, or clonidine.⁴⁷

Continue to: Once a patient is deemed...

Once a patient is deemed to be a candidate for stimulants, clinicians need to choose between methylphenidate or amphetamine/dextroamphetamine formulations. Table 6 lists medications that are commonly prescribed to treat ADHD; unless otherwise noted, these are FDA-approved for this indication. As a general rule, for adults, long-acting stimulant formulations are preferred over short-acting formulations.²⁸ Immediate-release stimulants are more prone to misuse or diversion compared to extended-release medications.²⁹ Longer-acting formulations may also provide better full-day symptom control.⁴⁸

In contrast to many other psychiatric medications, it may be beneficial to encourage periodically taking breaks or “medication holidays” from stimulants. Planned medication holidays for adults can involve intentionally not taking the medication over the weekend when the patient is not involved in work or school responsibilities. Such breaks have been shown to reduce adverse effects of stimulants (such as appetite suppression and insomnia) without significantly increasing ADHD symptoms.⁴⁹ Short breaks can also help prevent medication tolerance and the subsequent need to increase doses.⁵⁰ Medication holidays provide an opportunity to verify the ongoing benefits of the medication. It is advisable to periodically assess whether there is a continued need for stimulant treatment.⁵¹ If patients do not tolerate stimulants or have other contraindications, nonstimulants should be considered.

Lastly, no psychiatric patient should be treated with medication alone, and nonpharmacologic approaches should be incorporated as needed. Clear instructions, visual aids, nonverbal cues, frequent breaks to stand and stretch, schedules, normalizing failure as part of growth, and identifying triggers for emotional reactivity may help patients with ADHD.⁵² In a study of the academic performance of 92 college students taking medication for ADHD and 146 control students, treatment with stimulants alone did not eliminate the academic achievement deficit of those individuals with ADHD.⁵³ Good study habits (even without stimulants) appeared more important in overcoming the achievement disparity of students with ADHD.⁵³ Providing psychoeducation and training in concrete organization and planning skills have shown benefit.⁵⁴ Practice of skills on a daily basis appears to be especially beneficial.⁵⁵

Bottom Line

A sensible approach to diagnosing attention-deficit/hyperactivity disorder (ADHD) in adults includes ruling out other disorders that may present similar to ADHD, taking an appropriate developmental history, obtaining collateral information, and assessing for functional impairment. Sensible treatment involves ruling out comorbidities that stimulants could worsen, selecting extended-release stimulants, incorporating medication holidays, and using nonpharmacologic interventions.

Related Resources

• National Institute for Health and Care Excellence. Attention deficit hyperactivity disorder: diagnosis and management. https://www.nice.org.uk/guidance/ng87
• Substance Abuse and Mental Health Services Administration. Advisory: Prescription Stimulant Misuse Among Youth and Young Adults. https://store.samhsa.gov/product/prescription-stimulant-misuse-among-youth-young-adults/PEP21-06-01-003

Drug Brand Names

Amphetamine • Adzenys, Dyanavel, others
Atomoxetine • Strattera
Bupropion • Wellbutrin, Forfivo
Clonidine • Catapres, Kapvay
Dexmethylphenidate • Focalin
Dextroamphetamine • Dexedrine
Dextroamphetamine and amphetamine • Adderall, Mydayis
Guanfacine • Intuniv, Tenex
Lisdexamfetamine • Vyvanse
Methylphenidate • Concerta, Methylin, others
Viloxazine • Qelbree

Attention-deficit/hyperactivity disorder (ADHD) is common, with an estimated worldwide prevalence of 5.29% among children and adolescents and 2.5% among adults.¹ DSM-5-TR classifies ADHD as a neurodevelop­mental disorder, “a group of conditions with onset in the developmental period [that] typically manifest early in development, often before the child enters school.”² Because of the expectation that ADHD symptoms emerge early in development, the diagnostic criteria specify that symptoms must have been present prior to age 12 to qualify as ADHD. However, recent years have shown a significant increase in the number of patients being diagnosed with ADHD for the first time in adulthood. One study found that the diagnosis of ADHD among adults in the United States doubled between 2007 and 2016.³

First-line treatment for ADHD is the stimulants methylphenidate and amphetamine/dextroamphetamine. In the United States, these medications are classified as Schedule II controlled substances, indicating a high risk for abuse. However, just as ADHD diagnoses among adults have increased, so have prescriptions for stimulants. For example, Olfson et al⁴ found that stimulant prescriptions among young adults increased by a factor of 10 between 1994 and 2009.

The increased prevalence of adult patients diagnosed with ADHD and taking stimulants frequently places clinicians in a position to consider the validity of existing diagnoses and evaluate new patients with ADHD-related concerns. In this article, we review some of the challenges associated with diagnosing ADHD in adults, discuss the risks of stimulant treatment, and present a practical approach to the diagnosis and treatment of ADHD in adults.

Challenges in diagnosis

DSM-5-TR diagnostic criteria for ADHD are summarized in Table 1. Establishing a diagnosis of adult ADHD can be challenging. As with many psychiatric conditions, symptoms of ADHD are highly subjective. Retrospectively diagnosing a developmental condition in adults is often biased by the patient’s current functioning.⁵ ADHD has a high heritability and adults may inquire about the diagnosis if their children are diagnosed with ADHD.⁶ Some experts have cautioned that clinicians must be careful in diagnosing ADHD in adults.⁷ Just as there are risks associated with underdiagnosing ADHD, there are risks associated with overdiagnosis. Overdiagnosis may medicalize normal variants in the population and lead to unnecessary treatment and a misappropriation of limited medical resources.⁸ Many false positive cases of late-onset ADHD may be attributable to nonimpairing cognitive fluctuations.⁹

Poor diagnostic practices can impede accuracy in establishing the presence or absence of ADHD. Unfortunately, methods of diagnosing adult ADHD have been shown to vary widely in terms of information sources, diagnostic instruments used, symptom threshold, and whether functional impairment is a requirement for diagnosis.¹⁰ A common practice in diagnosing adult ADHD involves asking patients to complete self-report questionnaires that list symptoms of ADHD, such as the Adult ADHD Self-Report Scale developed by the World Health Organization.¹¹ However, self-reports of ADHD in adults are less reliable than informant reports, and some young adults without ADHD overreport symptoms.^12,13 Symptom checklists are particularly susceptible to faking, which lessens their diagnostic value.¹⁴

The possibility of malingered symptoms of ADHD further increases the diagnostic difficulty. College students may be particularly susceptible to overreporting ADHD symptoms in order to obtain academic accommodations or stimulants in the hopes of improving school performance.¹⁵ One study found that 25% to 48% of college students self-referred for ADHD evaluations exaggerated their symptoms.¹⁶ In another study, 31% of adults failed the Word Memory Test, which suggests noncredible performance in their ADHD evaluation.¹⁷ College students can successfully feign ADHD symptoms in both self-reported symptoms and computer-based tests of attention.¹⁸ Harrison et al¹⁹ summarized many of these concerns in their 2007 study of ADHD malingering, noting the “almost perfect ability of the Faking group to choose items … that correspond to the DSM-IV symptoms, and to report these at levels even higher than persons with diagnosed ADHD.” They suggested “Clinicians should be suspicious of students or young adults presenting for a first-time diagnosis who rate themselves as being significantly symptomatic, yet have managed to achieve well in school and other life activities.”¹⁹

Another challenge in correctly diagnosing adult ADHD is identifying other conditions that may impair attention.²⁰ Psychiatric conditions that may impair concentration include anxiety disorders, chronic stress, posttraumatic stress disorder, recent trauma, major depressive disorder (MDD), and bipolar disorder (BD). Undiagnosed learning disorders may present like ADHD. Focus can be negatively affected by sleep disorders such as sleep apnea, restless leg syndrome, or delayed sleep phase-onset disorder. Marijuana, cocaine, 3,4-methylenedioxy-methamphetamine (MDMA; “ecstasy”), caffeine, or prescription medications such as anticholinergics can also impair attention. Medical conditions that can present with attentional or executive functioning deficits include seizures, Lyme disease, HIV, encephalopathy, hypothyroidism, and “chemo brain.”²¹ Environmental factors such as age-related cognitive decline, sleep deprivation, inflammation, obesity, air pollution, chemical exposure, and excessive use of digital media may also produce symptoms similar to ADHD. Two studies of adult-onset ADHD concluded that 93% to 95% of cases were better explained by other conditions such as sleep disorders, substance use disorders, or another psychiatric disorder.²²

Continue to: Risks associated with treatment

Risks associated with treatment

With or without an accurate ADHD diagnosis, prescribing stimulants presents certain risks (Table 2^23-40). One of the more well-known risks of stimulants is addiction or misuse.²³ An estimated 5 million American adults misused prescription stimulants in 2016.²⁴ Despite stimulants’ status as controlled substances, long-term concurrent use of stimulants with opioids is common among adults with ADHD.²⁵ College students are particularly susceptible to misusing or diverting stimulants, often to improve their academic performance.²⁶ At 1 university, 22% of students had misused stimulants in the past year.²⁷ Prescribing short-acting stimulants (rather than extended-release formulations) increases the likelihood of misuse.²⁸ Patients prescribed stimulants begin to receive requests to divert their medications to others as early as elementary school, and by college more than one-third of those taking stimulants have been asked to give, sell, or trade their medications.²⁹ Diversion of stimulants by students with ADHD is prevalent, with 62% of patients engaging in diversion during their lifetime.¹⁵ Diverted stimulants can come from family members, black market sources, or deceived clinicians.³⁰ Although students’ stimulant misuse/diversion often is academically motivated, nonmedical use of psychostimulants does not appear to have a statistically significant effect on improving grade point average.³¹ Despite a negligible impact on grades, most students who take stimulants identify their effect as strongly positive, producing a situation in which misusers of stimulants have little motivation to stop.³² While some patients might ask for a stimulant prescription with the rationale that liking the effects proves they have ADHD, this is inappropriate because most individuals like the effects of stimulant medications.³³

The use of stimulants increases the risk for several adverse psychiatric outcomes. Stimulants increase the risk of anxiety, so exercise caution when prescribing to patients with a comorbid anxiety disorder.³⁴ Stimulants can also worsen irritability and insomnia, 2 issues common among patients with ADHD.³² Use of stimulant medications can trigger manic episodes. Viktorin et al³⁵ found a >6-fold increase in manic episodes among patients with BD receiving methylphenidate monotherapy compared to those receiving a combination of methyl­phenidate and a mood stabilizer.³⁵ The use of methylphenidate and amphetamine can lead to new-onset psychosis (or exacerbation of pre-existing psychotic illness); amphetamine use is associated with a higher risk of psychosis than methylphenidate.³⁶

General medical adverse effects are also possible with stimulant use. Stimulants’ adverse effect profiles include appetite suppression, dry mouth, and nausea. Long-term use poses a risk for stunting growth in children.¹ Using stimulants during pregnancy is associated with higher risk for neonatal morbidity, including preterm birth, CNS-related disorders, and seizures.³⁷ Stimulants can raise blood pressure and increase heart rate. Serious cardiovascular events associated with stimulant use include ventricular arrhythmias, strokes, and transient ischemic attacks.³⁸

Nonstimulant ADHD treatments are less risky than stimulants but still require monitoring for common adverse effects. Atomoxetine has been associated with sedation, growth retardation (in children), and in severe cases, liver injury or suicidal ideation.³⁹ Bupropion (commonly used off-label for ADHD) can lower the seizure threshold and cause irritability, anorexia, and insomnia.³⁹ Viloxazine, a newer agent, can cause hypertension, increased heart rate, nausea, drowsiness, headache, and insomnia.⁴⁰

Sensible diagnosing

Given the challenges in accurately diagnosing ADHD in adults, we present a sensible approach to making the diagnosis (Table 3). The first step is to rule out other conditions that might better explain the patient’s symptoms. A thorough clinical interview (including a psychiatric review of symptoms) is the cornerstone of an initial diagnostic assessment. The use of validated screening questionnaires such as the Patient Health Questionnaire-9 and General Anxiety Disorder-7 may also provide information regarding psychiatric conditions that require additional evaluation.

Continue to: Some of the most common conditions...

Some of the most common conditions we see mistaken for ADHD are MDD, generalized anxiety disorder (GAD), and BD. In DSM-5-TR, 1 of the diagnostic criteria for MDD is “diminished ability to think or concentrate, or indecisiveness, nearly every day (either by subjective account or as observed by others).”⁴¹ Similarly, criteria for GAD include “difficulty concentrating.”⁴² DSM-5-TR also includes distractibility as one of the criteria for mania/hypomania. Table 4^20-22,41,42 lists other psychiatric, substance-related, medical, and environmental conditions that can produce ADHD-like symptoms. Referring to some medical and environmental explanations for inattention, Aiken²² pointed out, “Patients who suffer from these problems might ask their doctor for a stimulant, but none of those syndromes require a psychopharmacologic approach.” ADHD can be comorbid with other psychiatric conditions, so the presence of another psychiatric illness does not automatically rule out ADHD. If alternative psychiatric diagnoses have been identified, these can be discussed with the patient and treatment offered that targets the specified condition.

Once alternative explanations have been ruled out, focus on the patient’s developmental history. DSM-5-TR conceptualizes ADHD as a neurodevelopmental disorder, meaning it is expected to emerge early in life. Whereas previous editions of DSM specified that ADHD symptoms must be present before age 7, DSM-5 modified this age threshold to before age 12.¹ This necessitates taking a careful life history in order to understand the presence or absence of symptoms at earlier developmental stages.⁵ ADHD should be verified by symptoms apparent in childhood and present across the lifespan.¹⁵

While this retrospective history is necessary, histories that rely on self-report alone are often unreliable. Collateral sources of information are generally more reliable when assessing for ADHD symptoms.¹³ Third-party sources can help confirm that any impairment is best attributed to ADHD rather than to another condition.¹⁵ Unfortunately, the difficulty of obtaining collateral information means it is often neglected, even in the literature.¹⁰ A parent is the ideal informant for gathering collateral information regarding a patient’s functioning in childhood.⁵ Suggested best practices also include obtaining collateral information from interviews with significant others, behavioral questionnaires completed by parents (for current and childhood symptoms), review of school records, and consideration of intellectual and achievement testing.⁴³ If psychological testing is pursued, include validity testing to detect feigned symptoms.^18,44

When evaluating for ADHD, assess not only for the presence of symptoms, but also if these symptoms produce significant functional impairment.^13,15 Impairments in daily functioning can include impaired school participation, social participation, quality of relationships, family conflict, family activities, family functioning, and emotional functioning.⁴⁵ Some symptoms may affect functioning in an adult’s life differently than they did during childhood, from missed work appointments to being late picking up kids from school. Research has shown that the correlation between the number of symptoms and functional impairment is weak, which means someone could experience all of the symptoms of ADHD without experiencing functional impairment.⁴⁵ To make an accurate diagnosis, it is therefore important to clearly establish both the number of symptoms the patient is experiencing and whether these symptoms are clearly linked to functional impairments.¹⁰

Sensible treatment

Once a diagnosis of ADHD has been clearly established, clinicians need to consider how best to treat the condition (Table 5). Stimulants are generally considered first-line treatment for ADHD. In randomized clinical trials, they showed significant efficacy; for example, one study of 146 adults with ADHD found a 76% improvement with methylphenidate compared to 19% for the placebo group.⁴⁶ Before starting a stimulant, certain comorbidities should be ruled out. If a patient has glaucoma or pheochromocytoma, they may first need treatment from or clearance by other specialists. Stimulants should likely be held in patients with hypertension, angina, or cardiovascular defects until receiving medical clearance. The risks of stimulants need to be discussed with female patients of childbearing age, weighing the benefits of treatment against the risks of medication use should the patient get pregnant. Patients with comorbid psychosis or uncontrolled bipolar illness should not receive stimulants due to the risk of exacerbation. Patients with active substance use disorders (SUDs) are generally not good candidates for stimulants because of the risk of misusing or diverting stimulants and the possibility that substance abuse may be causing their inattentive symptoms. Patients whose SUDs are in remission may cautiously be considered as candidates for stimulants. If patients misuse their prescribed stimulants, they should be switched to a nonstimulant medication such as atomoxetine, bupropion, guanfacine, or clonidine.⁴⁷

Continue to: Once a patient is deemed...

Once a patient is deemed to be a candidate for stimulants, clinicians need to choose between methylphenidate or amphetamine/dextroamphetamine formulations. Table 6 lists medications that are commonly prescribed to treat ADHD; unless otherwise noted, these are FDA-approved for this indication. As a general rule, for adults, long-acting stimulant formulations are preferred over short-acting formulations.²⁸ Immediate-release stimulants are more prone to misuse or diversion compared to extended-release medications.²⁹ Longer-acting formulations may also provide better full-day symptom control.⁴⁸

In contrast to many other psychiatric medications, it may be beneficial to encourage periodically taking breaks or “medication holidays” from stimulants. Planned medication holidays for adults can involve intentionally not taking the medication over the weekend when the patient is not involved in work or school responsibilities. Such breaks have been shown to reduce adverse effects of stimulants (such as appetite suppression and insomnia) without significantly increasing ADHD symptoms.⁴⁹ Short breaks can also help prevent medication tolerance and the subsequent need to increase doses.⁵⁰ Medication holidays provide an opportunity to verify the ongoing benefits of the medication. It is advisable to periodically assess whether there is a continued need for stimulant treatment.⁵¹ If patients do not tolerate stimulants or have other contraindications, nonstimulants should be considered.

Lastly, no psychiatric patient should be treated with medication alone, and nonpharmacologic approaches should be incorporated as needed. Clear instructions, visual aids, nonverbal cues, frequent breaks to stand and stretch, schedules, normalizing failure as part of growth, and identifying triggers for emotional reactivity may help patients with ADHD.⁵² In a study of the academic performance of 92 college students taking medication for ADHD and 146 control students, treatment with stimulants alone did not eliminate the academic achievement deficit of those individuals with ADHD.⁵³ Good study habits (even without stimulants) appeared more important in overcoming the achievement disparity of students with ADHD.⁵³ Providing psychoeducation and training in concrete organization and planning skills have shown benefit.⁵⁴ Practice of skills on a daily basis appears to be especially beneficial.⁵⁵

Bottom Line

A sensible approach to diagnosing attention-deficit/hyperactivity disorder (ADHD) in adults includes ruling out other disorders that may present similar to ADHD, taking an appropriate developmental history, obtaining collateral information, and assessing for functional impairment. Sensible treatment involves ruling out comorbidities that stimulants could worsen, selecting extended-release stimulants, incorporating medication holidays, and using nonpharmacologic interventions.

Related Resources

• National Institute for Health and Care Excellence. Attention deficit hyperactivity disorder: diagnosis and management. https://www.nice.org.uk/guidance/ng87
• Substance Abuse and Mental Health Services Administration. Advisory: Prescription Stimulant Misuse Among Youth and Young Adults. https://store.samhsa.gov/product/prescription-stimulant-misuse-among-youth-young-adults/PEP21-06-01-003

Drug Brand Names

Amphetamine • Adzenys, Dyanavel, others
Atomoxetine • Strattera
Bupropion • Wellbutrin, Forfivo
Clonidine • Catapres, Kapvay
Dexmethylphenidate • Focalin
Dextroamphetamine • Dexedrine
Dextroamphetamine and amphetamine • Adderall, Mydayis
Guanfacine • Intuniv, Tenex
Lisdexamfetamine • Vyvanse
Methylphenidate • Concerta, Methylin, others
Viloxazine • Qelbree

Attention-deficit/hyperactivity disorder (ADHD) is common, with an estimated worldwide prevalence of 5.29% among children and adolescents and 2.5% among adults.¹ DSM-5-TR classifies ADHD as a neurodevelop­mental disorder, “a group of conditions with onset in the developmental period [that] typically manifest early in development, often before the child enters school.”² Because of the expectation that ADHD symptoms emerge early in development, the diagnostic criteria specify that symptoms must have been present prior to age 12 to qualify as ADHD. However, recent years have shown a significant increase in the number of patients being diagnosed with ADHD for the first time in adulthood. One study found that the diagnosis of ADHD among adults in the United States doubled between 2007 and 2016.³

First-line treatment for ADHD is the stimulants methylphenidate and amphetamine/dextroamphetamine. In the United States, these medications are classified as Schedule II controlled substances, indicating a high risk for abuse. However, just as ADHD diagnoses among adults have increased, so have prescriptions for stimulants. For example, Olfson et al⁴ found that stimulant prescriptions among young adults increased by a factor of 10 between 1994 and 2009.

The increased prevalence of adult patients diagnosed with ADHD and taking stimulants frequently places clinicians in a position to consider the validity of existing diagnoses and evaluate new patients with ADHD-related concerns. In this article, we review some of the challenges associated with diagnosing ADHD in adults, discuss the risks of stimulant treatment, and present a practical approach to the diagnosis and treatment of ADHD in adults.

Challenges in diagnosis

DSM-5-TR diagnostic criteria for ADHD are summarized in Table 1. Establishing a diagnosis of adult ADHD can be challenging. As with many psychiatric conditions, symptoms of ADHD are highly subjective. Retrospectively diagnosing a developmental condition in adults is often biased by the patient’s current functioning.⁵ ADHD has a high heritability and adults may inquire about the diagnosis if their children are diagnosed with ADHD.⁶ Some experts have cautioned that clinicians must be careful in diagnosing ADHD in adults.⁷ Just as there are risks associated with underdiagnosing ADHD, there are risks associated with overdiagnosis. Overdiagnosis may medicalize normal variants in the population and lead to unnecessary treatment and a misappropriation of limited medical resources.⁸ Many false positive cases of late-onset ADHD may be attributable to nonimpairing cognitive fluctuations.⁹

Poor diagnostic practices can impede accuracy in establishing the presence or absence of ADHD. Unfortunately, methods of diagnosing adult ADHD have been shown to vary widely in terms of information sources, diagnostic instruments used, symptom threshold, and whether functional impairment is a requirement for diagnosis.¹⁰ A common practice in diagnosing adult ADHD involves asking patients to complete self-report questionnaires that list symptoms of ADHD, such as the Adult ADHD Self-Report Scale developed by the World Health Organization.¹¹ However, self-reports of ADHD in adults are less reliable than informant reports, and some young adults without ADHD overreport symptoms.^12,13 Symptom checklists are particularly susceptible to faking, which lessens their diagnostic value.¹⁴

The possibility of malingered symptoms of ADHD further increases the diagnostic difficulty. College students may be particularly susceptible to overreporting ADHD symptoms in order to obtain academic accommodations or stimulants in the hopes of improving school performance.¹⁵ One study found that 25% to 48% of college students self-referred for ADHD evaluations exaggerated their symptoms.¹⁶ In another study, 31% of adults failed the Word Memory Test, which suggests noncredible performance in their ADHD evaluation.¹⁷ College students can successfully feign ADHD symptoms in both self-reported symptoms and computer-based tests of attention.¹⁸ Harrison et al¹⁹ summarized many of these concerns in their 2007 study of ADHD malingering, noting the “almost perfect ability of the Faking group to choose items … that correspond to the DSM-IV symptoms, and to report these at levels even higher than persons with diagnosed ADHD.” They suggested “Clinicians should be suspicious of students or young adults presenting for a first-time diagnosis who rate themselves as being significantly symptomatic, yet have managed to achieve well in school and other life activities.”¹⁹

Another challenge in correctly diagnosing adult ADHD is identifying other conditions that may impair attention.²⁰ Psychiatric conditions that may impair concentration include anxiety disorders, chronic stress, posttraumatic stress disorder, recent trauma, major depressive disorder (MDD), and bipolar disorder (BD). Undiagnosed learning disorders may present like ADHD. Focus can be negatively affected by sleep disorders such as sleep apnea, restless leg syndrome, or delayed sleep phase-onset disorder. Marijuana, cocaine, 3,4-methylenedioxy-methamphetamine (MDMA; “ecstasy”), caffeine, or prescription medications such as anticholinergics can also impair attention. Medical conditions that can present with attentional or executive functioning deficits include seizures, Lyme disease, HIV, encephalopathy, hypothyroidism, and “chemo brain.”²¹ Environmental factors such as age-related cognitive decline, sleep deprivation, inflammation, obesity, air pollution, chemical exposure, and excessive use of digital media may also produce symptoms similar to ADHD. Two studies of adult-onset ADHD concluded that 93% to 95% of cases were better explained by other conditions such as sleep disorders, substance use disorders, or another psychiatric disorder.²²

Continue to: Risks associated with treatment

Risks associated with treatment

With or without an accurate ADHD diagnosis, prescribing stimulants presents certain risks (Table 2^23-40). One of the more well-known risks of stimulants is addiction or misuse.²³ An estimated 5 million American adults misused prescription stimulants in 2016.²⁴ Despite stimulants’ status as controlled substances, long-term concurrent use of stimulants with opioids is common among adults with ADHD.²⁵ College students are particularly susceptible to misusing or diverting stimulants, often to improve their academic performance.²⁶ At 1 university, 22% of students had misused stimulants in the past year.²⁷ Prescribing short-acting stimulants (rather than extended-release formulations) increases the likelihood of misuse.²⁸ Patients prescribed stimulants begin to receive requests to divert their medications to others as early as elementary school, and by college more than one-third of those taking stimulants have been asked to give, sell, or trade their medications.²⁹ Diversion of stimulants by students with ADHD is prevalent, with 62% of patients engaging in diversion during their lifetime.¹⁵ Diverted stimulants can come from family members, black market sources, or deceived clinicians.³⁰ Although students’ stimulant misuse/diversion often is academically motivated, nonmedical use of psychostimulants does not appear to have a statistically significant effect on improving grade point average.³¹ Despite a negligible impact on grades, most students who take stimulants identify their effect as strongly positive, producing a situation in which misusers of stimulants have little motivation to stop.³² While some patients might ask for a stimulant prescription with the rationale that liking the effects proves they have ADHD, this is inappropriate because most individuals like the effects of stimulant medications.³³

The use of stimulants increases the risk for several adverse psychiatric outcomes. Stimulants increase the risk of anxiety, so exercise caution when prescribing to patients with a comorbid anxiety disorder.³⁴ Stimulants can also worsen irritability and insomnia, 2 issues common among patients with ADHD.³² Use of stimulant medications can trigger manic episodes. Viktorin et al³⁵ found a >6-fold increase in manic episodes among patients with BD receiving methylphenidate monotherapy compared to those receiving a combination of methyl­phenidate and a mood stabilizer.³⁵ The use of methylphenidate and amphetamine can lead to new-onset psychosis (or exacerbation of pre-existing psychotic illness); amphetamine use is associated with a higher risk of psychosis than methylphenidate.³⁶

General medical adverse effects are also possible with stimulant use. Stimulants’ adverse effect profiles include appetite suppression, dry mouth, and nausea. Long-term use poses a risk for stunting growth in children.¹ Using stimulants during pregnancy is associated with higher risk for neonatal morbidity, including preterm birth, CNS-related disorders, and seizures.³⁷ Stimulants can raise blood pressure and increase heart rate. Serious cardiovascular events associated with stimulant use include ventricular arrhythmias, strokes, and transient ischemic attacks.³⁸

Nonstimulant ADHD treatments are less risky than stimulants but still require monitoring for common adverse effects. Atomoxetine has been associated with sedation, growth retardation (in children), and in severe cases, liver injury or suicidal ideation.³⁹ Bupropion (commonly used off-label for ADHD) can lower the seizure threshold and cause irritability, anorexia, and insomnia.³⁹ Viloxazine, a newer agent, can cause hypertension, increased heart rate, nausea, drowsiness, headache, and insomnia.⁴⁰

Sensible diagnosing

Given the challenges in accurately diagnosing ADHD in adults, we present a sensible approach to making the diagnosis (Table 3). The first step is to rule out other conditions that might better explain the patient’s symptoms. A thorough clinical interview (including a psychiatric review of symptoms) is the cornerstone of an initial diagnostic assessment. The use of validated screening questionnaires such as the Patient Health Questionnaire-9 and General Anxiety Disorder-7 may also provide information regarding psychiatric conditions that require additional evaluation.

Continue to: Some of the most common conditions...

Some of the most common conditions we see mistaken for ADHD are MDD, generalized anxiety disorder (GAD), and BD. In DSM-5-TR, 1 of the diagnostic criteria for MDD is “diminished ability to think or concentrate, or indecisiveness, nearly every day (either by subjective account or as observed by others).”⁴¹ Similarly, criteria for GAD include “difficulty concentrating.”⁴² DSM-5-TR also includes distractibility as one of the criteria for mania/hypomania. Table 4^20-22,41,42 lists other psychiatric, substance-related, medical, and environmental conditions that can produce ADHD-like symptoms. Referring to some medical and environmental explanations for inattention, Aiken²² pointed out, “Patients who suffer from these problems might ask their doctor for a stimulant, but none of those syndromes require a psychopharmacologic approach.” ADHD can be comorbid with other psychiatric conditions, so the presence of another psychiatric illness does not automatically rule out ADHD. If alternative psychiatric diagnoses have been identified, these can be discussed with the patient and treatment offered that targets the specified condition.

Once alternative explanations have been ruled out, focus on the patient’s developmental history. DSM-5-TR conceptualizes ADHD as a neurodevelopmental disorder, meaning it is expected to emerge early in life. Whereas previous editions of DSM specified that ADHD symptoms must be present before age 7, DSM-5 modified this age threshold to before age 12.¹ This necessitates taking a careful life history in order to understand the presence or absence of symptoms at earlier developmental stages.⁵ ADHD should be verified by symptoms apparent in childhood and present across the lifespan.¹⁵

While this retrospective history is necessary, histories that rely on self-report alone are often unreliable. Collateral sources of information are generally more reliable when assessing for ADHD symptoms.¹³ Third-party sources can help confirm that any impairment is best attributed to ADHD rather than to another condition.¹⁵ Unfortunately, the difficulty of obtaining collateral information means it is often neglected, even in the literature.¹⁰ A parent is the ideal informant for gathering collateral information regarding a patient’s functioning in childhood.⁵ Suggested best practices also include obtaining collateral information from interviews with significant others, behavioral questionnaires completed by parents (for current and childhood symptoms), review of school records, and consideration of intellectual and achievement testing.⁴³ If psychological testing is pursued, include validity testing to detect feigned symptoms.^18,44

When evaluating for ADHD, assess not only for the presence of symptoms, but also if these symptoms produce significant functional impairment.^13,15 Impairments in daily functioning can include impaired school participation, social participation, quality of relationships, family conflict, family activities, family functioning, and emotional functioning.⁴⁵ Some symptoms may affect functioning in an adult’s life differently than they did during childhood, from missed work appointments to being late picking up kids from school. Research has shown that the correlation between the number of symptoms and functional impairment is weak, which means someone could experience all of the symptoms of ADHD without experiencing functional impairment.⁴⁵ To make an accurate diagnosis, it is therefore important to clearly establish both the number of symptoms the patient is experiencing and whether these symptoms are clearly linked to functional impairments.¹⁰

Sensible treatment

Once a diagnosis of ADHD has been clearly established, clinicians need to consider how best to treat the condition (Table 5). Stimulants are generally considered first-line treatment for ADHD. In randomized clinical trials, they showed significant efficacy; for example, one study of 146 adults with ADHD found a 76% improvement with methylphenidate compared to 19% for the placebo group.⁴⁶ Before starting a stimulant, certain comorbidities should be ruled out. If a patient has glaucoma or pheochromocytoma, they may first need treatment from or clearance by other specialists. Stimulants should likely be held in patients with hypertension, angina, or cardiovascular defects until receiving medical clearance. The risks of stimulants need to be discussed with female patients of childbearing age, weighing the benefits of treatment against the risks of medication use should the patient get pregnant. Patients with comorbid psychosis or uncontrolled bipolar illness should not receive stimulants due to the risk of exacerbation. Patients with active substance use disorders (SUDs) are generally not good candidates for stimulants because of the risk of misusing or diverting stimulants and the possibility that substance abuse may be causing their inattentive symptoms. Patients whose SUDs are in remission may cautiously be considered as candidates for stimulants. If patients misuse their prescribed stimulants, they should be switched to a nonstimulant medication such as atomoxetine, bupropion, guanfacine, or clonidine.⁴⁷

Continue to: Once a patient is deemed...

Once a patient is deemed to be a candidate for stimulants, clinicians need to choose between methylphenidate or amphetamine/dextroamphetamine formulations. Table 6 lists medications that are commonly prescribed to treat ADHD; unless otherwise noted, these are FDA-approved for this indication. As a general rule, for adults, long-acting stimulant formulations are preferred over short-acting formulations.²⁸ Immediate-release stimulants are more prone to misuse or diversion compared to extended-release medications.²⁹ Longer-acting formulations may also provide better full-day symptom control.⁴⁸

In contrast to many other psychiatric medications, it may be beneficial to encourage periodically taking breaks or “medication holidays” from stimulants. Planned medication holidays for adults can involve intentionally not taking the medication over the weekend when the patient is not involved in work or school responsibilities. Such breaks have been shown to reduce adverse effects of stimulants (such as appetite suppression and insomnia) without significantly increasing ADHD symptoms.⁴⁹ Short breaks can also help prevent medication tolerance and the subsequent need to increase doses.⁵⁰ Medication holidays provide an opportunity to verify the ongoing benefits of the medication. It is advisable to periodically assess whether there is a continued need for stimulant treatment.⁵¹ If patients do not tolerate stimulants or have other contraindications, nonstimulants should be considered.

Lastly, no psychiatric patient should be treated with medication alone, and nonpharmacologic approaches should be incorporated as needed. Clear instructions, visual aids, nonverbal cues, frequent breaks to stand and stretch, schedules, normalizing failure as part of growth, and identifying triggers for emotional reactivity may help patients with ADHD.⁵² In a study of the academic performance of 92 college students taking medication for ADHD and 146 control students, treatment with stimulants alone did not eliminate the academic achievement deficit of those individuals with ADHD.⁵³ Good study habits (even without stimulants) appeared more important in overcoming the achievement disparity of students with ADHD.⁵³ Providing psychoeducation and training in concrete organization and planning skills have shown benefit.⁵⁴ Practice of skills on a daily basis appears to be especially beneficial.⁵⁵

Bottom Line

A sensible approach to diagnosing attention-deficit/hyperactivity disorder (ADHD) in adults includes ruling out other disorders that may present similar to ADHD, taking an appropriate developmental history, obtaining collateral information, and assessing for functional impairment. Sensible treatment involves ruling out comorbidities that stimulants could worsen, selecting extended-release stimulants, incorporating medication holidays, and using nonpharmacologic interventions.

Related Resources

• National Institute for Health and Care Excellence. Attention deficit hyperactivity disorder: diagnosis and management. https://www.nice.org.uk/guidance/ng87
• Substance Abuse and Mental Health Services Administration. Advisory: Prescription Stimulant Misuse Among Youth and Young Adults. https://store.samhsa.gov/product/prescription-stimulant-misuse-among-youth-young-adults/PEP21-06-01-003

Drug Brand Names

Amphetamine • Adzenys, Dyanavel, others
Atomoxetine • Strattera
Bupropion • Wellbutrin, Forfivo
Clonidine • Catapres, Kapvay
Dexmethylphenidate • Focalin
Dextroamphetamine • Dexedrine
Dextroamphetamine and amphetamine • Adderall, Mydayis
Guanfacine • Intuniv, Tenex
Lisdexamfetamine • Vyvanse
Methylphenidate • Concerta, Methylin, others
Viloxazine • Qelbree

Patients with chronic insomnia that does not improve with nonpharmacologic techniques often develop tolerance to sedative medications (benzodiazepines) prescribed for nightly use. When nonbenzodiazepine medications are used, tachyphylaxis can develop and these medications no longer initiate or maintain sleep. Strategies that alternate between these 2 types of agents are simple to follow and may allow patients to maintain sensitivity to both types of medications. In this article, I review the types, causes, evaluation, and treatment of insomnia; describe an alternating medication strategy to help patients avoid developing tolerance/tachyphylaxis; and present 3 fictional case vignettes to illustrate this approach.

A common, troubling condition

Insomnia is a common problem among psychiatric patients. Approximately 30% to 50% of adults experience occasional, short-term (<3 months) insomnia, and 5% to 10% experience chronic (≥3 months) insomnia,¹ with associated negative impacts on health and quality of life. Insomnia is sometimes primary and may have a hereditary component, but more often is associated with medical, neurologic, or psychiatric disorders.

Patterns of insomnia include difficulty falling asleep (initial or sleep-onset insomnia), remaining asleep (middle or sleep-maintenance insomnia), or falling back asleep after early awakening (late or sleep-offset insomnia). Sleep-onset insomnia correlates with high levels of anxiety and worrying, but once asleep, patients usually stay asleep. Sleep-maintenance problems involve multiple awakenings after falling asleep and taking hours to fall back to sleep. These patients experience inadequate sleep when they must wake up early for school or work. Early-awakening patients report feeling wide awake by 4 to 5 am and being unable to get back to sleep.

Caffeine is an important consideration for patients with sleep difficulties. Its use is widespread in much of the world, whether ingested as coffee, tea, in soft drinks, or in “energy” drinks that may contain as much as 200 mg of caffeine (twice the amount in a typical cup of brewed coffee). Caffeine may also be ingested as an ingredient of medications for headache or migraine. While some individuals maintain that they can fall asleep easily after drinking caffeinated coffee, many may not recognize the amount of caffeine they consume and its negative impact on sleep.² Author Michael Pollan stopped use of all caffeine and reported on the surprising positive effect on his sleep.³

Patients with mood, anxiety, or psychotic disorders are likely to experience insomnia intermittently or chronically, and insomnia predisposes some individuals to develop mood and anxiety symptoms.⁴ Patients with insomnia often experience anxiety focused on a fear of not getting adequate sleep, which creates a vicious cycle in which hyperarousal associated with fear of not sleeping complicates other causes of insomnia. A patient’s chronotype (preference for the time of day in which they carry out activities vs sleeping) also may play a role in sleep difficulties (Box⁵).

Box

Certain medications may contribute to insomnia, particularly stimulants. It is important to understand and explain to patients the time frame during which immediate-release or extended-release (ER) stimulants are active, which varies in individuals depending on liver enzyme activity. Other commonly used psychotropic medications—including bupropion, modafinil, armodafinil, atomoxetine, amphetamine salts, and methylphenidate—may interfere with sleep if used later in the day.⁶

Patients typically do not mention their use of alcohol and/or marijuana unless asked. Those who are binge drinkers or alcohol-dependent may expect alcohol to help them fall asleep, but usually find their sleep is disrupted and difficult to maintain. Patients may use marijuana to help them sleep, particularly marijuana high in tetrahydrocannabinol (THC). While it may help with sleep initiation, THC can disrupt sleep maintenance. Cannabidiol does not have intrinsic sedating effects and may even interfere with sleep.^7,8

Continue to: Women may be more likely...

Women may be more likely than men to experience insomnia.⁹ The onset of menopause can bring hot flashes that interfere with sleep.

Women with a history of mood disorders are more likely to have a history of premenstrual dysphoric disorder, postpartum depression, and unusual responses to oral contraceptives.¹⁰ These women are more likely to report problems with mood, energy, and sleep at perimenopause. Treatment with estrogen replacement may be an option for women without risk factors, such as clotting disorders, smoking history, or a personal or family history of breast or uterine cancer. For many who are not candidates for or who refuse estrogen replacement, use of a selective serotonin reuptake inhibitor (SSRI) or serotonin-norepinephrine reuptake inhibitor at low doses may help with vasomotor symptoms but not with insomnia.

Insomnia symptoms typically increase with age.¹¹ When sleep is adequate early in life but becomes a problem in midlife, an individual’s eating habits, obesity, and lack of exercise may be contributing factors. The typical American diet includes highly refined carbohydrates with excess salt; such foods are often readily available to the exclusion of healthy options. Overweight and obese patients may insist they eat a healthy diet with 3 meals per day, but a careful history often uncovers nighttime binge eating. Nighttime binge eating is rarely reported. This not only maintains obesity, but also interferes with sleep, since patients stay up late to avoid discovery by family members.¹² This lack of sleep can lead to an endless loop because insufficient sleep is a risk factor for obesity.¹³

Evaluating sleep difficulties

New patient evaluations should include a careful history beginning with childhood, including personal early childhood history and family psychiatric history. Patients often report the onset of sleep difficulty and anxiety during childhood, which should raise further questions about aspects of mood regulation from early life such as concentration, energy, motivation, appetite, and academic performance. While many children and adolescents are diagnosed with attention-deficit/hyperactivity disorder due to concentration problems that cause difficulties at school, be aware this might be part of a syndrome related to mood regulation.¹⁴ Unexpected responses to an SSRI—such as agitation, euphoria, or an immediate response with the first dose—should also raise suspicion of a mood disorder. Once the underlying mood disorder is stabilized, many patients report improved sleep.¹⁵

If a patient reports having difficulty falling and remaining asleep but is not sure if there is a pattern, keeping a sleep diary can help. Further questioning may uncover the cause. Does the patient have spontaneous jerks of lower extremities (restless leg syndrome) that interfere with falling asleep or wake them up? Have they noticed problems with dreams/nightmares that wake them, which could be associated with posttraumatic stress, anxiety, or depression? Have they been told by a partner that they act out dreams or are seemingly awake but not responsive, which could point to REM sleep behavior disorder or early Parkinson’s disease? Referral to a sleep laboratory and a neurologist can help establish the correct diagnosis and point to appropriate treatment.^16-18

Treatment options

Several cognitive-behavioral techniques, including cognitive-behavioral therapy for insomnia (CBT-I), yogic breathing, progressive relaxation, mindfulness meditation, and sleep hygiene techniques may help considerably,^19,20 but insomnia often remains difficult to treat. Pharmacotherapy is not necessarily more effective than nonpharmacologic approaches. Both options require the patient to take initiative to either find nonpharmacologic approaches or discuss the problem with a physician and agree to take medication.²¹ A trial comparing CBT-I to sedatives or the combination of CBT-I plus sedatives found higher rates of sleep with CBT-I for 3 months, after which improvement fluctuated; the combination showed sustained improvement for the entire 6-month trial.²² CBT-I has also been shown to be as effective with patients who do not have psychiatric illness as for those who are depressed, anxious, or stressed.²³ However, behavioral techniques that require regular practice may be difficult for individuals to maintain, particularly when they are depressed or anxious.

Continue to: Clinicians should understand...

Clinicians should understand the distinctions among the various types of pharmacotherapy for insomnia. Sedative-hypnotics include medications with varying half-lives and metabolic pathways. Short-acting benzodiazepines such as triazolam or alprazolam and the “z-drugs” zolpidem or zaleplon may help initiate sleep in patients with sleep-onset insomnia. Longer-acting benzodiazepines such as diazepam, clonazepam, or temazepam and the z-drug eszopiclone may also help with sleep maintenance.²³ Based on my clinical experience, individual patients may respond better to 1 type of medication over another, or even to different agents within the same class of sedative-hypnotics.

Some clinicians prescribe nonbenzodiazepine medications for sleep, such as doxepin (which is FDA-approved for treating insomnia) or off-label trazodone, mirtazapine, or quetiapine. Their antihistaminic properties confer sedating effects. Virtually all over-the-counter (OTC) medications for insomnia are antihistaminic. These OTC medications are not designed to treat insomnia, and the optimal dosage to maintain sleep without daytime sedation must be determined by trial and error. Sedating nonbenzodiazepine medications may be slowly absorbed if taken at bedtime (depending on whether they are taken with or without food) and cause daytime sedation and cognitive slowness in patients with sleep-onset and maintenance insomnia who must wake up early. Starting trazodone at 50 to 75 mg may cause slow metabolizers to wake up with considerable sedation, while fast metabolizers might never feel soundly asleep.²⁴

Patients with mood and anxiety disorders that complicate insomnia are often prescribed second-generation antipsychotics such as quetiapine, lurasidone, or olanzapine, which are sedating as well as mood-stabilizing. These approaches require careful attention to titrating doses and timing their use.

Problems with pharmacotherapy

When either benzodiazepines or nonbenzodiazepine medications are used on a long-standing, nightly basis, they often stop working well. It is not unusual that after days to weeks of taking a benzodiazepine, patients find they no longer stay asleep but can’t fall asleep if they don’t take them. Once tolerance develops, the individual experiences pharmacologic withdrawal with an inability to fall asleep or stay asleep. The medication becomes necessary but ineffective, and many patients increase their use to higher doses to fall asleep, and sometimes in early morning to maintain sleep. This leads to negative effects on cognition, coordination/balance, and mood during the day, especially in older patients.

Nonbenzodiazepine sedating medications do not lead to pharmacologic tolerance but do lead to tachyphylaxis as the CNS attempts to downregulate sedation to keep the organism safe. For some patients, this happens quickly, within a matter of days.²⁵ Others increase doses to stay asleep. For example, a patient with a starting dose of trazodone 75 mg/d might increase the dosage to 300 mg/d. While trazodone is approved in doses of 300 to 600 mg as an antidepressant, it is preferable to keep doses lower when used only for sedation.

Continue to: An alternating medication strategy

Alternating between medications from different classes can help patients avoid developing tolerance with benzodiazepines or tachyphylaxis as occurs with antihistaminic medications. It can be effective for patients with primary insomnia as well as for those whose sleep problems are associated with mood or anxiety disorders. Patients typically maintain sensitivity to any form of pharmacologic sedation for several nights without loss of effect but need to take a break to maintain the sedation effect. For example, in 1 case study, a 30-year-old female who rapidly developed tachyphylaxis to the sedative action of mirtazapine experienced a return of the medication’s sedative effects after taking a 3-day break.²⁵

To initiate an alternating strategy, the clinician must first help the patient establish a sedating dose of 2 medications from different classes, such as trazodone and zolpidem, and then instruct the patient to use each for 2 to 3 consecutive nights on an alternating basis. Patients can use calendars or pillboxes to avoid confusion about which medication to take on a given night. In many cases, this approach can work indefinitely.

The following 3 case vignettes illustrate how this alternating medication strategy can work.

CASE 1

Mr. B, age 58, is a married salesman whose territory includes 3 states. He drives from client to client from Monday through Thursday each week, staying overnight in hotels. He is comfortable talking to clients, has a close and supportive relationship with his wife, and enjoys socializing with friends. Mr. B has a high level of trait anxiety and perfectionism and is proud of his sales record throughout his career, but this leads to insomnia during his nights on the road, and often on Sunday night as he starts anticipating the week ahead. Mr. B denies having a depressed mood or cognitive problems. When on vacation with his wife he has no trouble sleeping. He has no psychiatric family history or any substantial medical problems. He simply wishes that he could sleep on work nights.

We set up an alternating medication approach. Mr. B takes trazodone 100 mg on the first night and 150 mg on the second and third nights. He then takes triazolam 0.25 mg for 2 nights; previously, he had found that zolpidem did not work as well for maintaining sleep. He can sleep adequately for the 2 weekend nights, then restarts the alternating pattern. Mr. B has done well with this regimen for >10 years.

Continue to: CASE 2

Ms. C, age 60, is widowed and has a successful career as a corporate attorney. She has been anxious since early childhood and has had trouble falling asleep for much of her life. Once she falls asleep on her sofa—often between 1 and 2 am—Ms. C can sleep soundly for 7 to 8 hours, but early morning work meetings require her to set an alarm for 6 am daily. Ms. C feels irritable and anxious on awakening but arrives at her office by 7:30 am, where she maintains a full schedule, with frequent 12-hour workdays. Ms. C did not experience significant insomnia or hot flashes with menopause at age 52 and does not use hormone replacement therapy.

Ms. C denies having depression, but experienced appropriate grief related to her husband’s illness and death from metastatic cancer 3 years ago. At the time, her internist prescribed escitalopram and zolpidem; escitalopram caused greater agitation and distress, so she stopped it after 10 days. Zolpidem 10 mg/d allowed her to sleep but she worried about taking it because her mother had long-standing sedative dependence. Ms. C lives alone, but her adult children live nearby, and she has a strong support system that includes colleagues at her firm, friends at her book club, and a support group for partners of cancer patients.

Ms. C tries trazodone, starting with 50 mg, but reports feeling agitated rather than sleepy and has cognitive fogginess in the morning. She is switched to quetiapine 50 mg, which she tolerates well and allows her to sleep soundly. To avoid developing tachyphylaxis with quetiapine, she takes eszopiclone 3 mg for 2 nights, alternating with quetiapine for 3 nights. This strategy allows her to reliably fall asleep by 11 pm, wake up at 6 am, and feel rested throughout the day.

CASE 3

Ms. D, age 55, is married with a long-standing diagnosis of generalized anxiety disorder (GAD), panic disorder, and depression so severe she is unable to work as a preschool teacher. She notes that past clinicians have prescribed a wide array of antidepressants and benzodiazepines but she remains anxious, agitated, and unable to sleep. She worries constantly about running out of benzodiazepines, which are “the only medication that helps me.” At the time of evaluation, her medications are venlafaxine ER 150 mg/d, lorazepam 1 mg 3 times daily and 2 mg at bedtime, and buspirone 15 mg 3 times daily, which she admits to not taking. She is overweight and does not exercise. She spends her days snacking and watching television. She can’t settle down enough to read and feels overwhelmed most of the time. Her adult children won’t allow her to babysit their young children because she dozes during the day.

Ms. D has a strong family history of psychiatric illness, including a father with bipolar I disorder and alcohol use disorder and a sister with schizoaffective disorder. Ms. D has never felt overtly manic, but has spent most of her life feeling depressed, anxious, and hopeless, and at times she has wished she was dead. She has had poor responses to many antidepressants, with transient euphoria followed by more anxiety.

Continue to: Rather than major depressive disorder...

Rather than major depressive disorder or GAD, Ms. D’s symptoms better meet the criteria for bipolar II disorder. She agrees to a slow taper of venlafaxine and a slow increase of divalproex, starting with 125 mg each evening. While taking venlafaxine 75 mg/d and divalproex 375 mg/d, she experiences distinct improvement in anxiety and agitation, which further improve after venlafaxine is stopped and divalproex is increased to 750 mg in the evening. She finds that she forgets daytime doses of lorazepam but depends on it to fall asleep. While taking quetiapine 50 mg and lorazepam 1 mg at bedtime, Ms. D reports sleeping soundly and feeling alert in the morning. Over several weeks, she tapers lorazepam slowly by 0.5 mg every 2 weeks. She finds she needs a higher dose of quetiapine to stay asleep, eventually requiring 400 mg each night. Ms. D says overall she feels better but is distressed because she has gained 25 lbs since starting divalproex and quetiapine.

To avoid further increases in quetiapine and maintain its sedating effect, Ms. D is switched to an alternating schedule of clonazepam 1.5 mg for 2 nights and quetiapine 300 mg for 3 nights. She agrees to begin exercising by walking in her neighborhood daily, and gradually increases this to 1 hour per day. After starting to exercise regularly, she finds she is oversedated by quetiapine at night, so she is gradually decreased to a dose of 150 mg, while still alternating with clonazepam 1.5 mg. Ms. D loses most of the weight she had gained and begins volunteering as a reading coach in the elementary school in her neighborhood.

Bottom Line

Patients with chronic insomnia can often maintain adequate sedation without developing tolerance to benzodiazepines or tachyphylaxis with nonsedating agents by using 2 sleep medications that have different mechanisms of action on an alternating schedule.

Related Resources

• Sateia MJ, Buysse DJ, Krystal AD, et al. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2): 307-349. doi:10.5664/jcsm.6470
• Muppavarapu K, Muthukanagaraj M, Saeed SA. Cognitive-behavioral therapy for insomnia: a review of 8 studies. Current Psychiatry. 2020;19(9):40-46. doi:10.12788/cp.0040

Drug Brand Names

Alprazolam • Xanax
Armodafinil • Nuvigil
Atomoxetine • Strattera
Bupropion • Wellbutrin
Clonazepam • Klonopin
Diazepam • Valium
Divalproex • Depakote
Doxepin • Sinequan
Escitalopram • Lexapro
Eszopiclone • Lunesta
Lorazepam • Ativan
Lurasidone • Latuda
Methylphenidate • Concerta
Mirtazapine • Remeron
Modafinil • Provigil
Olanzapine • Zyprexa
Quetiapine • Seroquel
Temazepam • Restoril
Trazodone • Desyrel
Triazolam • Halcion
Venlafaxine • Effexor
Zaleplon • Sonata
Zolpidem • Ambien

Patients with chronic insomnia that does not improve with nonpharmacologic techniques often develop tolerance to sedative medications (benzodiazepines) prescribed for nightly use. When nonbenzodiazepine medications are used, tachyphylaxis can develop and these medications no longer initiate or maintain sleep. Strategies that alternate between these 2 types of agents are simple to follow and may allow patients to maintain sensitivity to both types of medications. In this article, I review the types, causes, evaluation, and treatment of insomnia; describe an alternating medication strategy to help patients avoid developing tolerance/tachyphylaxis; and present 3 fictional case vignettes to illustrate this approach.

A common, troubling condition

Insomnia is a common problem among psychiatric patients. Approximately 30% to 50% of adults experience occasional, short-term (<3 months) insomnia, and 5% to 10% experience chronic (≥3 months) insomnia,¹ with associated negative impacts on health and quality of life. Insomnia is sometimes primary and may have a hereditary component, but more often is associated with medical, neurologic, or psychiatric disorders.

Patterns of insomnia include difficulty falling asleep (initial or sleep-onset insomnia), remaining asleep (middle or sleep-maintenance insomnia), or falling back asleep after early awakening (late or sleep-offset insomnia). Sleep-onset insomnia correlates with high levels of anxiety and worrying, but once asleep, patients usually stay asleep. Sleep-maintenance problems involve multiple awakenings after falling asleep and taking hours to fall back to sleep. These patients experience inadequate sleep when they must wake up early for school or work. Early-awakening patients report feeling wide awake by 4 to 5 am and being unable to get back to sleep.

Caffeine is an important consideration for patients with sleep difficulties. Its use is widespread in much of the world, whether ingested as coffee, tea, in soft drinks, or in “energy” drinks that may contain as much as 200 mg of caffeine (twice the amount in a typical cup of brewed coffee). Caffeine may also be ingested as an ingredient of medications for headache or migraine. While some individuals maintain that they can fall asleep easily after drinking caffeinated coffee, many may not recognize the amount of caffeine they consume and its negative impact on sleep.² Author Michael Pollan stopped use of all caffeine and reported on the surprising positive effect on his sleep.³

Patients with mood, anxiety, or psychotic disorders are likely to experience insomnia intermittently or chronically, and insomnia predisposes some individuals to develop mood and anxiety symptoms.⁴ Patients with insomnia often experience anxiety focused on a fear of not getting adequate sleep, which creates a vicious cycle in which hyperarousal associated with fear of not sleeping complicates other causes of insomnia. A patient’s chronotype (preference for the time of day in which they carry out activities vs sleeping) also may play a role in sleep difficulties (Box⁵).

Box

Certain medications may contribute to insomnia, particularly stimulants. It is important to understand and explain to patients the time frame during which immediate-release or extended-release (ER) stimulants are active, which varies in individuals depending on liver enzyme activity. Other commonly used psychotropic medications—including bupropion, modafinil, armodafinil, atomoxetine, amphetamine salts, and methylphenidate—may interfere with sleep if used later in the day.⁶

Patients typically do not mention their use of alcohol and/or marijuana unless asked. Those who are binge drinkers or alcohol-dependent may expect alcohol to help them fall asleep, but usually find their sleep is disrupted and difficult to maintain. Patients may use marijuana to help them sleep, particularly marijuana high in tetrahydrocannabinol (THC). While it may help with sleep initiation, THC can disrupt sleep maintenance. Cannabidiol does not have intrinsic sedating effects and may even interfere with sleep.^7,8

Continue to: Women may be more likely...

Women may be more likely than men to experience insomnia.⁹ The onset of menopause can bring hot flashes that interfere with sleep.

Women with a history of mood disorders are more likely to have a history of premenstrual dysphoric disorder, postpartum depression, and unusual responses to oral contraceptives.¹⁰ These women are more likely to report problems with mood, energy, and sleep at perimenopause. Treatment with estrogen replacement may be an option for women without risk factors, such as clotting disorders, smoking history, or a personal or family history of breast or uterine cancer. For many who are not candidates for or who refuse estrogen replacement, use of a selective serotonin reuptake inhibitor (SSRI) or serotonin-norepinephrine reuptake inhibitor at low doses may help with vasomotor symptoms but not with insomnia.

Insomnia symptoms typically increase with age.¹¹ When sleep is adequate early in life but becomes a problem in midlife, an individual’s eating habits, obesity, and lack of exercise may be contributing factors. The typical American diet includes highly refined carbohydrates with excess salt; such foods are often readily available to the exclusion of healthy options. Overweight and obese patients may insist they eat a healthy diet with 3 meals per day, but a careful history often uncovers nighttime binge eating. Nighttime binge eating is rarely reported. This not only maintains obesity, but also interferes with sleep, since patients stay up late to avoid discovery by family members.¹² This lack of sleep can lead to an endless loop because insufficient sleep is a risk factor for obesity.¹³

Evaluating sleep difficulties

New patient evaluations should include a careful history beginning with childhood, including personal early childhood history and family psychiatric history. Patients often report the onset of sleep difficulty and anxiety during childhood, which should raise further questions about aspects of mood regulation from early life such as concentration, energy, motivation, appetite, and academic performance. While many children and adolescents are diagnosed with attention-deficit/hyperactivity disorder due to concentration problems that cause difficulties at school, be aware this might be part of a syndrome related to mood regulation.¹⁴ Unexpected responses to an SSRI—such as agitation, euphoria, or an immediate response with the first dose—should also raise suspicion of a mood disorder. Once the underlying mood disorder is stabilized, many patients report improved sleep.¹⁵

If a patient reports having difficulty falling and remaining asleep but is not sure if there is a pattern, keeping a sleep diary can help. Further questioning may uncover the cause. Does the patient have spontaneous jerks of lower extremities (restless leg syndrome) that interfere with falling asleep or wake them up? Have they noticed problems with dreams/nightmares that wake them, which could be associated with posttraumatic stress, anxiety, or depression? Have they been told by a partner that they act out dreams or are seemingly awake but not responsive, which could point to REM sleep behavior disorder or early Parkinson’s disease? Referral to a sleep laboratory and a neurologist can help establish the correct diagnosis and point to appropriate treatment.^16-18

Treatment options

Several cognitive-behavioral techniques, including cognitive-behavioral therapy for insomnia (CBT-I), yogic breathing, progressive relaxation, mindfulness meditation, and sleep hygiene techniques may help considerably,^19,20 but insomnia often remains difficult to treat. Pharmacotherapy is not necessarily more effective than nonpharmacologic approaches. Both options require the patient to take initiative to either find nonpharmacologic approaches or discuss the problem with a physician and agree to take medication.²¹ A trial comparing CBT-I to sedatives or the combination of CBT-I plus sedatives found higher rates of sleep with CBT-I for 3 months, after which improvement fluctuated; the combination showed sustained improvement for the entire 6-month trial.²² CBT-I has also been shown to be as effective with patients who do not have psychiatric illness as for those who are depressed, anxious, or stressed.²³ However, behavioral techniques that require regular practice may be difficult for individuals to maintain, particularly when they are depressed or anxious.

Continue to: Clinicians should understand...

Clinicians should understand the distinctions among the various types of pharmacotherapy for insomnia. Sedative-hypnotics include medications with varying half-lives and metabolic pathways. Short-acting benzodiazepines such as triazolam or alprazolam and the “z-drugs” zolpidem or zaleplon may help initiate sleep in patients with sleep-onset insomnia. Longer-acting benzodiazepines such as diazepam, clonazepam, or temazepam and the z-drug eszopiclone may also help with sleep maintenance.²³ Based on my clinical experience, individual patients may respond better to 1 type of medication over another, or even to different agents within the same class of sedative-hypnotics.

Some clinicians prescribe nonbenzodiazepine medications for sleep, such as doxepin (which is FDA-approved for treating insomnia) or off-label trazodone, mirtazapine, or quetiapine. Their antihistaminic properties confer sedating effects. Virtually all over-the-counter (OTC) medications for insomnia are antihistaminic. These OTC medications are not designed to treat insomnia, and the optimal dosage to maintain sleep without daytime sedation must be determined by trial and error. Sedating nonbenzodiazepine medications may be slowly absorbed if taken at bedtime (depending on whether they are taken with or without food) and cause daytime sedation and cognitive slowness in patients with sleep-onset and maintenance insomnia who must wake up early. Starting trazodone at 50 to 75 mg may cause slow metabolizers to wake up with considerable sedation, while fast metabolizers might never feel soundly asleep.²⁴

Patients with mood and anxiety disorders that complicate insomnia are often prescribed second-generation antipsychotics such as quetiapine, lurasidone, or olanzapine, which are sedating as well as mood-stabilizing. These approaches require careful attention to titrating doses and timing their use.

Problems with pharmacotherapy

When either benzodiazepines or nonbenzodiazepine medications are used on a long-standing, nightly basis, they often stop working well. It is not unusual that after days to weeks of taking a benzodiazepine, patients find they no longer stay asleep but can’t fall asleep if they don’t take them. Once tolerance develops, the individual experiences pharmacologic withdrawal with an inability to fall asleep or stay asleep. The medication becomes necessary but ineffective, and many patients increase their use to higher doses to fall asleep, and sometimes in early morning to maintain sleep. This leads to negative effects on cognition, coordination/balance, and mood during the day, especially in older patients.

Nonbenzodiazepine sedating medications do not lead to pharmacologic tolerance but do lead to tachyphylaxis as the CNS attempts to downregulate sedation to keep the organism safe. For some patients, this happens quickly, within a matter of days.²⁵ Others increase doses to stay asleep. For example, a patient with a starting dose of trazodone 75 mg/d might increase the dosage to 300 mg/d. While trazodone is approved in doses of 300 to 600 mg as an antidepressant, it is preferable to keep doses lower when used only for sedation.

Continue to: An alternating medication strategy

Alternating between medications from different classes can help patients avoid developing tolerance with benzodiazepines or tachyphylaxis as occurs with antihistaminic medications. It can be effective for patients with primary insomnia as well as for those whose sleep problems are associated with mood or anxiety disorders. Patients typically maintain sensitivity to any form of pharmacologic sedation for several nights without loss of effect but need to take a break to maintain the sedation effect. For example, in 1 case study, a 30-year-old female who rapidly developed tachyphylaxis to the sedative action of mirtazapine experienced a return of the medication’s sedative effects after taking a 3-day break.²⁵

To initiate an alternating strategy, the clinician must first help the patient establish a sedating dose of 2 medications from different classes, such as trazodone and zolpidem, and then instruct the patient to use each for 2 to 3 consecutive nights on an alternating basis. Patients can use calendars or pillboxes to avoid confusion about which medication to take on a given night. In many cases, this approach can work indefinitely.

The following 3 case vignettes illustrate how this alternating medication strategy can work.

CASE 1

Mr. B, age 58, is a married salesman whose territory includes 3 states. He drives from client to client from Monday through Thursday each week, staying overnight in hotels. He is comfortable talking to clients, has a close and supportive relationship with his wife, and enjoys socializing with friends. Mr. B has a high level of trait anxiety and perfectionism and is proud of his sales record throughout his career, but this leads to insomnia during his nights on the road, and often on Sunday night as he starts anticipating the week ahead. Mr. B denies having a depressed mood or cognitive problems. When on vacation with his wife he has no trouble sleeping. He has no psychiatric family history or any substantial medical problems. He simply wishes that he could sleep on work nights.

We set up an alternating medication approach. Mr. B takes trazodone 100 mg on the first night and 150 mg on the second and third nights. He then takes triazolam 0.25 mg for 2 nights; previously, he had found that zolpidem did not work as well for maintaining sleep. He can sleep adequately for the 2 weekend nights, then restarts the alternating pattern. Mr. B has done well with this regimen for >10 years.

Continue to: CASE 2

Ms. C, age 60, is widowed and has a successful career as a corporate attorney. She has been anxious since early childhood and has had trouble falling asleep for much of her life. Once she falls asleep on her sofa—often between 1 and 2 am—Ms. C can sleep soundly for 7 to 8 hours, but early morning work meetings require her to set an alarm for 6 am daily. Ms. C feels irritable and anxious on awakening but arrives at her office by 7:30 am, where she maintains a full schedule, with frequent 12-hour workdays. Ms. C did not experience significant insomnia or hot flashes with menopause at age 52 and does not use hormone replacement therapy.

Ms. C denies having depression, but experienced appropriate grief related to her husband’s illness and death from metastatic cancer 3 years ago. At the time, her internist prescribed escitalopram and zolpidem; escitalopram caused greater agitation and distress, so she stopped it after 10 days. Zolpidem 10 mg/d allowed her to sleep but she worried about taking it because her mother had long-standing sedative dependence. Ms. C lives alone, but her adult children live nearby, and she has a strong support system that includes colleagues at her firm, friends at her book club, and a support group for partners of cancer patients.

Ms. C tries trazodone, starting with 50 mg, but reports feeling agitated rather than sleepy and has cognitive fogginess in the morning. She is switched to quetiapine 50 mg, which she tolerates well and allows her to sleep soundly. To avoid developing tachyphylaxis with quetiapine, she takes eszopiclone 3 mg for 2 nights, alternating with quetiapine for 3 nights. This strategy allows her to reliably fall asleep by 11 pm, wake up at 6 am, and feel rested throughout the day.

CASE 3

Ms. D, age 55, is married with a long-standing diagnosis of generalized anxiety disorder (GAD), panic disorder, and depression so severe she is unable to work as a preschool teacher. She notes that past clinicians have prescribed a wide array of antidepressants and benzodiazepines but she remains anxious, agitated, and unable to sleep. She worries constantly about running out of benzodiazepines, which are “the only medication that helps me.” At the time of evaluation, her medications are venlafaxine ER 150 mg/d, lorazepam 1 mg 3 times daily and 2 mg at bedtime, and buspirone 15 mg 3 times daily, which she admits to not taking. She is overweight and does not exercise. She spends her days snacking and watching television. She can’t settle down enough to read and feels overwhelmed most of the time. Her adult children won’t allow her to babysit their young children because she dozes during the day.

Ms. D has a strong family history of psychiatric illness, including a father with bipolar I disorder and alcohol use disorder and a sister with schizoaffective disorder. Ms. D has never felt overtly manic, but has spent most of her life feeling depressed, anxious, and hopeless, and at times she has wished she was dead. She has had poor responses to many antidepressants, with transient euphoria followed by more anxiety.

Continue to: Rather than major depressive disorder...

Rather than major depressive disorder or GAD, Ms. D’s symptoms better meet the criteria for bipolar II disorder. She agrees to a slow taper of venlafaxine and a slow increase of divalproex, starting with 125 mg each evening. While taking venlafaxine 75 mg/d and divalproex 375 mg/d, she experiences distinct improvement in anxiety and agitation, which further improve after venlafaxine is stopped and divalproex is increased to 750 mg in the evening. She finds that she forgets daytime doses of lorazepam but depends on it to fall asleep. While taking quetiapine 50 mg and lorazepam 1 mg at bedtime, Ms. D reports sleeping soundly and feeling alert in the morning. Over several weeks, she tapers lorazepam slowly by 0.5 mg every 2 weeks. She finds she needs a higher dose of quetiapine to stay asleep, eventually requiring 400 mg each night. Ms. D says overall she feels better but is distressed because she has gained 25 lbs since starting divalproex and quetiapine.

To avoid further increases in quetiapine and maintain its sedating effect, Ms. D is switched to an alternating schedule of clonazepam 1.5 mg for 2 nights and quetiapine 300 mg for 3 nights. She agrees to begin exercising by walking in her neighborhood daily, and gradually increases this to 1 hour per day. After starting to exercise regularly, she finds she is oversedated by quetiapine at night, so she is gradually decreased to a dose of 150 mg, while still alternating with clonazepam 1.5 mg. Ms. D loses most of the weight she had gained and begins volunteering as a reading coach in the elementary school in her neighborhood.

Bottom Line

Patients with chronic insomnia can often maintain adequate sedation without developing tolerance to benzodiazepines or tachyphylaxis with nonsedating agents by using 2 sleep medications that have different mechanisms of action on an alternating schedule.

Related Resources

• Sateia MJ, Buysse DJ, Krystal AD, et al. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2): 307-349. doi:10.5664/jcsm.6470
• Muppavarapu K, Muthukanagaraj M, Saeed SA. Cognitive-behavioral therapy for insomnia: a review of 8 studies. Current Psychiatry. 2020;19(9):40-46. doi:10.12788/cp.0040

Drug Brand Names

Alprazolam • Xanax
Armodafinil • Nuvigil
Atomoxetine • Strattera
Bupropion • Wellbutrin
Clonazepam • Klonopin
Diazepam • Valium
Divalproex • Depakote
Doxepin • Sinequan
Escitalopram • Lexapro
Eszopiclone • Lunesta
Lorazepam • Ativan
Lurasidone • Latuda
Methylphenidate • Concerta
Mirtazapine • Remeron
Modafinil • Provigil
Olanzapine • Zyprexa
Quetiapine • Seroquel
Temazepam • Restoril
Trazodone • Desyrel
Triazolam • Halcion
Venlafaxine • Effexor
Zaleplon • Sonata
Zolpidem • Ambien

Patients with chronic insomnia that does not improve with nonpharmacologic techniques often develop tolerance to sedative medications (benzodiazepines) prescribed for nightly use. When nonbenzodiazepine medications are used, tachyphylaxis can develop and these medications no longer initiate or maintain sleep. Strategies that alternate between these 2 types of agents are simple to follow and may allow patients to maintain sensitivity to both types of medications. In this article, I review the types, causes, evaluation, and treatment of insomnia; describe an alternating medication strategy to help patients avoid developing tolerance/tachyphylaxis; and present 3 fictional case vignettes to illustrate this approach.

A common, troubling condition

Insomnia is a common problem among psychiatric patients. Approximately 30% to 50% of adults experience occasional, short-term (<3 months) insomnia, and 5% to 10% experience chronic (≥3 months) insomnia,¹ with associated negative impacts on health and quality of life. Insomnia is sometimes primary and may have a hereditary component, but more often is associated with medical, neurologic, or psychiatric disorders.

Patterns of insomnia include difficulty falling asleep (initial or sleep-onset insomnia), remaining asleep (middle or sleep-maintenance insomnia), or falling back asleep after early awakening (late or sleep-offset insomnia). Sleep-onset insomnia correlates with high levels of anxiety and worrying, but once asleep, patients usually stay asleep. Sleep-maintenance problems involve multiple awakenings after falling asleep and taking hours to fall back to sleep. These patients experience inadequate sleep when they must wake up early for school or work. Early-awakening patients report feeling wide awake by 4 to 5 am and being unable to get back to sleep.

Caffeine is an important consideration for patients with sleep difficulties. Its use is widespread in much of the world, whether ingested as coffee, tea, in soft drinks, or in “energy” drinks that may contain as much as 200 mg of caffeine (twice the amount in a typical cup of brewed coffee). Caffeine may also be ingested as an ingredient of medications for headache or migraine. While some individuals maintain that they can fall asleep easily after drinking caffeinated coffee, many may not recognize the amount of caffeine they consume and its negative impact on sleep.² Author Michael Pollan stopped use of all caffeine and reported on the surprising positive effect on his sleep.³

Patients with mood, anxiety, or psychotic disorders are likely to experience insomnia intermittently or chronically, and insomnia predisposes some individuals to develop mood and anxiety symptoms.⁴ Patients with insomnia often experience anxiety focused on a fear of not getting adequate sleep, which creates a vicious cycle in which hyperarousal associated with fear of not sleeping complicates other causes of insomnia. A patient’s chronotype (preference for the time of day in which they carry out activities vs sleeping) also may play a role in sleep difficulties (Box⁵).

Box

Certain medications may contribute to insomnia, particularly stimulants. It is important to understand and explain to patients the time frame during which immediate-release or extended-release (ER) stimulants are active, which varies in individuals depending on liver enzyme activity. Other commonly used psychotropic medications—including bupropion, modafinil, armodafinil, atomoxetine, amphetamine salts, and methylphenidate—may interfere with sleep if used later in the day.⁶

Patients typically do not mention their use of alcohol and/or marijuana unless asked. Those who are binge drinkers or alcohol-dependent may expect alcohol to help them fall asleep, but usually find their sleep is disrupted and difficult to maintain. Patients may use marijuana to help them sleep, particularly marijuana high in tetrahydrocannabinol (THC). While it may help with sleep initiation, THC can disrupt sleep maintenance. Cannabidiol does not have intrinsic sedating effects and may even interfere with sleep.^7,8

Continue to: Women may be more likely...

Women may be more likely than men to experience insomnia.⁹ The onset of menopause can bring hot flashes that interfere with sleep.

Women with a history of mood disorders are more likely to have a history of premenstrual dysphoric disorder, postpartum depression, and unusual responses to oral contraceptives.¹⁰ These women are more likely to report problems with mood, energy, and sleep at perimenopause. Treatment with estrogen replacement may be an option for women without risk factors, such as clotting disorders, smoking history, or a personal or family history of breast or uterine cancer. For many who are not candidates for or who refuse estrogen replacement, use of a selective serotonin reuptake inhibitor (SSRI) or serotonin-norepinephrine reuptake inhibitor at low doses may help with vasomotor symptoms but not with insomnia.

Insomnia symptoms typically increase with age.¹¹ When sleep is adequate early in life but becomes a problem in midlife, an individual’s eating habits, obesity, and lack of exercise may be contributing factors. The typical American diet includes highly refined carbohydrates with excess salt; such foods are often readily available to the exclusion of healthy options. Overweight and obese patients may insist they eat a healthy diet with 3 meals per day, but a careful history often uncovers nighttime binge eating. Nighttime binge eating is rarely reported. This not only maintains obesity, but also interferes with sleep, since patients stay up late to avoid discovery by family members.¹² This lack of sleep can lead to an endless loop because insufficient sleep is a risk factor for obesity.¹³

Evaluating sleep difficulties

New patient evaluations should include a careful history beginning with childhood, including personal early childhood history and family psychiatric history. Patients often report the onset of sleep difficulty and anxiety during childhood, which should raise further questions about aspects of mood regulation from early life such as concentration, energy, motivation, appetite, and academic performance. While many children and adolescents are diagnosed with attention-deficit/hyperactivity disorder due to concentration problems that cause difficulties at school, be aware this might be part of a syndrome related to mood regulation.¹⁴ Unexpected responses to an SSRI—such as agitation, euphoria, or an immediate response with the first dose—should also raise suspicion of a mood disorder. Once the underlying mood disorder is stabilized, many patients report improved sleep.¹⁵

If a patient reports having difficulty falling and remaining asleep but is not sure if there is a pattern, keeping a sleep diary can help. Further questioning may uncover the cause. Does the patient have spontaneous jerks of lower extremities (restless leg syndrome) that interfere with falling asleep or wake them up? Have they noticed problems with dreams/nightmares that wake them, which could be associated with posttraumatic stress, anxiety, or depression? Have they been told by a partner that they act out dreams or are seemingly awake but not responsive, which could point to REM sleep behavior disorder or early Parkinson’s disease? Referral to a sleep laboratory and a neurologist can help establish the correct diagnosis and point to appropriate treatment.^16-18

Treatment options

Several cognitive-behavioral techniques, including cognitive-behavioral therapy for insomnia (CBT-I), yogic breathing, progressive relaxation, mindfulness meditation, and sleep hygiene techniques may help considerably,^19,20 but insomnia often remains difficult to treat. Pharmacotherapy is not necessarily more effective than nonpharmacologic approaches. Both options require the patient to take initiative to either find nonpharmacologic approaches or discuss the problem with a physician and agree to take medication.²¹ A trial comparing CBT-I to sedatives or the combination of CBT-I plus sedatives found higher rates of sleep with CBT-I for 3 months, after which improvement fluctuated; the combination showed sustained improvement for the entire 6-month trial.²² CBT-I has also been shown to be as effective with patients who do not have psychiatric illness as for those who are depressed, anxious, or stressed.²³ However, behavioral techniques that require regular practice may be difficult for individuals to maintain, particularly when they are depressed or anxious.

Continue to: Clinicians should understand...

Clinicians should understand the distinctions among the various types of pharmacotherapy for insomnia. Sedative-hypnotics include medications with varying half-lives and metabolic pathways. Short-acting benzodiazepines such as triazolam or alprazolam and the “z-drugs” zolpidem or zaleplon may help initiate sleep in patients with sleep-onset insomnia. Longer-acting benzodiazepines such as diazepam, clonazepam, or temazepam and the z-drug eszopiclone may also help with sleep maintenance.²³ Based on my clinical experience, individual patients may respond better to 1 type of medication over another, or even to different agents within the same class of sedative-hypnotics.

Some clinicians prescribe nonbenzodiazepine medications for sleep, such as doxepin (which is FDA-approved for treating insomnia) or off-label trazodone, mirtazapine, or quetiapine. Their antihistaminic properties confer sedating effects. Virtually all over-the-counter (OTC) medications for insomnia are antihistaminic. These OTC medications are not designed to treat insomnia, and the optimal dosage to maintain sleep without daytime sedation must be determined by trial and error. Sedating nonbenzodiazepine medications may be slowly absorbed if taken at bedtime (depending on whether they are taken with or without food) and cause daytime sedation and cognitive slowness in patients with sleep-onset and maintenance insomnia who must wake up early. Starting trazodone at 50 to 75 mg may cause slow metabolizers to wake up with considerable sedation, while fast metabolizers might never feel soundly asleep.²⁴

Patients with mood and anxiety disorders that complicate insomnia are often prescribed second-generation antipsychotics such as quetiapine, lurasidone, or olanzapine, which are sedating as well as mood-stabilizing. These approaches require careful attention to titrating doses and timing their use.

Problems with pharmacotherapy

When either benzodiazepines or nonbenzodiazepine medications are used on a long-standing, nightly basis, they often stop working well. It is not unusual that after days to weeks of taking a benzodiazepine, patients find they no longer stay asleep but can’t fall asleep if they don’t take them. Once tolerance develops, the individual experiences pharmacologic withdrawal with an inability to fall asleep or stay asleep. The medication becomes necessary but ineffective, and many patients increase their use to higher doses to fall asleep, and sometimes in early morning to maintain sleep. This leads to negative effects on cognition, coordination/balance, and mood during the day, especially in older patients.

Nonbenzodiazepine sedating medications do not lead to pharmacologic tolerance but do lead to tachyphylaxis as the CNS attempts to downregulate sedation to keep the organism safe. For some patients, this happens quickly, within a matter of days.²⁵ Others increase doses to stay asleep. For example, a patient with a starting dose of trazodone 75 mg/d might increase the dosage to 300 mg/d. While trazodone is approved in doses of 300 to 600 mg as an antidepressant, it is preferable to keep doses lower when used only for sedation.

Continue to: An alternating medication strategy

Alternating between medications from different classes can help patients avoid developing tolerance with benzodiazepines or tachyphylaxis as occurs with antihistaminic medications. It can be effective for patients with primary insomnia as well as for those whose sleep problems are associated with mood or anxiety disorders. Patients typically maintain sensitivity to any form of pharmacologic sedation for several nights without loss of effect but need to take a break to maintain the sedation effect. For example, in 1 case study, a 30-year-old female who rapidly developed tachyphylaxis to the sedative action of mirtazapine experienced a return of the medication’s sedative effects after taking a 3-day break.²⁵

To initiate an alternating strategy, the clinician must first help the patient establish a sedating dose of 2 medications from different classes, such as trazodone and zolpidem, and then instruct the patient to use each for 2 to 3 consecutive nights on an alternating basis. Patients can use calendars or pillboxes to avoid confusion about which medication to take on a given night. In many cases, this approach can work indefinitely.

The following 3 case vignettes illustrate how this alternating medication strategy can work.

CASE 1

Mr. B, age 58, is a married salesman whose territory includes 3 states. He drives from client to client from Monday through Thursday each week, staying overnight in hotels. He is comfortable talking to clients, has a close and supportive relationship with his wife, and enjoys socializing with friends. Mr. B has a high level of trait anxiety and perfectionism and is proud of his sales record throughout his career, but this leads to insomnia during his nights on the road, and often on Sunday night as he starts anticipating the week ahead. Mr. B denies having a depressed mood or cognitive problems. When on vacation with his wife he has no trouble sleeping. He has no psychiatric family history or any substantial medical problems. He simply wishes that he could sleep on work nights.

We set up an alternating medication approach. Mr. B takes trazodone 100 mg on the first night and 150 mg on the second and third nights. He then takes triazolam 0.25 mg for 2 nights; previously, he had found that zolpidem did not work as well for maintaining sleep. He can sleep adequately for the 2 weekend nights, then restarts the alternating pattern. Mr. B has done well with this regimen for >10 years.

Continue to: CASE 2

Ms. C, age 60, is widowed and has a successful career as a corporate attorney. She has been anxious since early childhood and has had trouble falling asleep for much of her life. Once she falls asleep on her sofa—often between 1 and 2 am—Ms. C can sleep soundly for 7 to 8 hours, but early morning work meetings require her to set an alarm for 6 am daily. Ms. C feels irritable and anxious on awakening but arrives at her office by 7:30 am, where she maintains a full schedule, with frequent 12-hour workdays. Ms. C did not experience significant insomnia or hot flashes with menopause at age 52 and does not use hormone replacement therapy.

Ms. C denies having depression, but experienced appropriate grief related to her husband’s illness and death from metastatic cancer 3 years ago. At the time, her internist prescribed escitalopram and zolpidem; escitalopram caused greater agitation and distress, so she stopped it after 10 days. Zolpidem 10 mg/d allowed her to sleep but she worried about taking it because her mother had long-standing sedative dependence. Ms. C lives alone, but her adult children live nearby, and she has a strong support system that includes colleagues at her firm, friends at her book club, and a support group for partners of cancer patients.

Ms. C tries trazodone, starting with 50 mg, but reports feeling agitated rather than sleepy and has cognitive fogginess in the morning. She is switched to quetiapine 50 mg, which she tolerates well and allows her to sleep soundly. To avoid developing tachyphylaxis with quetiapine, she takes eszopiclone 3 mg for 2 nights, alternating with quetiapine for 3 nights. This strategy allows her to reliably fall asleep by 11 pm, wake up at 6 am, and feel rested throughout the day.

CASE 3

Ms. D, age 55, is married with a long-standing diagnosis of generalized anxiety disorder (GAD), panic disorder, and depression so severe she is unable to work as a preschool teacher. She notes that past clinicians have prescribed a wide array of antidepressants and benzodiazepines but she remains anxious, agitated, and unable to sleep. She worries constantly about running out of benzodiazepines, which are “the only medication that helps me.” At the time of evaluation, her medications are venlafaxine ER 150 mg/d, lorazepam 1 mg 3 times daily and 2 mg at bedtime, and buspirone 15 mg 3 times daily, which she admits to not taking. She is overweight and does not exercise. She spends her days snacking and watching television. She can’t settle down enough to read and feels overwhelmed most of the time. Her adult children won’t allow her to babysit their young children because she dozes during the day.

Ms. D has a strong family history of psychiatric illness, including a father with bipolar I disorder and alcohol use disorder and a sister with schizoaffective disorder. Ms. D has never felt overtly manic, but has spent most of her life feeling depressed, anxious, and hopeless, and at times she has wished she was dead. She has had poor responses to many antidepressants, with transient euphoria followed by more anxiety.

Continue to: Rather than major depressive disorder...

Rather than major depressive disorder or GAD, Ms. D’s symptoms better meet the criteria for bipolar II disorder. She agrees to a slow taper of venlafaxine and a slow increase of divalproex, starting with 125 mg each evening. While taking venlafaxine 75 mg/d and divalproex 375 mg/d, she experiences distinct improvement in anxiety and agitation, which further improve after venlafaxine is stopped and divalproex is increased to 750 mg in the evening. She finds that she forgets daytime doses of lorazepam but depends on it to fall asleep. While taking quetiapine 50 mg and lorazepam 1 mg at bedtime, Ms. D reports sleeping soundly and feeling alert in the morning. Over several weeks, she tapers lorazepam slowly by 0.5 mg every 2 weeks. She finds she needs a higher dose of quetiapine to stay asleep, eventually requiring 400 mg each night. Ms. D says overall she feels better but is distressed because she has gained 25 lbs since starting divalproex and quetiapine.

To avoid further increases in quetiapine and maintain its sedating effect, Ms. D is switched to an alternating schedule of clonazepam 1.5 mg for 2 nights and quetiapine 300 mg for 3 nights. She agrees to begin exercising by walking in her neighborhood daily, and gradually increases this to 1 hour per day. After starting to exercise regularly, she finds she is oversedated by quetiapine at night, so she is gradually decreased to a dose of 150 mg, while still alternating with clonazepam 1.5 mg. Ms. D loses most of the weight she had gained and begins volunteering as a reading coach in the elementary school in her neighborhood.

Bottom Line

Patients with chronic insomnia can often maintain adequate sedation without developing tolerance to benzodiazepines or tachyphylaxis with nonsedating agents by using 2 sleep medications that have different mechanisms of action on an alternating schedule.

Related Resources

• Sateia MJ, Buysse DJ, Krystal AD, et al. Clinical practice guideline for the pharmacologic treatment of chronic insomnia in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(2): 307-349. doi:10.5664/jcsm.6470
• Muppavarapu K, Muthukanagaraj M, Saeed SA. Cognitive-behavioral therapy for insomnia: a review of 8 studies. Current Psychiatry. 2020;19(9):40-46. doi:10.12788/cp.0040

Drug Brand Names

Alprazolam • Xanax
Armodafinil • Nuvigil
Atomoxetine • Strattera
Bupropion • Wellbutrin
Clonazepam • Klonopin
Diazepam • Valium
Divalproex • Depakote
Doxepin • Sinequan
Escitalopram • Lexapro
Eszopiclone • Lunesta
Lorazepam • Ativan
Lurasidone • Latuda
Methylphenidate • Concerta
Mirtazapine • Remeron
Modafinil • Provigil
Olanzapine • Zyprexa
Quetiapine • Seroquel
Temazepam • Restoril
Trazodone • Desyrel
Triazolam • Halcion
Venlafaxine • Effexor
Zaleplon • Sonata
Zolpidem • Ambien