Unfortunately, recent research has not unveiled a breakthrough for preventing or treating cognitive impairment or Alzheimer disease. But several studies from the last 2 years are helping to drive the field of geriatrics forward, providing evidence of what does and does not help a variety of issues specific to the elderly. 

Based on a search of the 2017 and 2018 literature, this article presents new evidence on preventing and treating cognitive impairment, managing dementia-associated behavioral disturbances and delirium, preventing falls, and improving inpatient mobility and posthospital care transitions.

COGNITIVE IMPAIRMENT, DEMENTIA: STILL NO SILVER BULLET

With the exception of oral anticoagulation treatment for atrial fibrillation, there is little evidence that pharmacologic or nonpharmacologic interventions slow the onset or progression of Alzheimer disease.

Nonpharmacologic interventions

Home occupational therapy. A 2-year home-based occupational therapy intervention¹ showed no evidence of slowing functional decline in patients with Alzheimer disease. The randomized controlled trial involving 180 participants consisted of monthly sessions of an intensive, well-established collaborative-care management model that included fall prevention and other safety strategies, personalized training in activities of daily living, exercise, and education. Outcome measures for activities of daily living did not differ significantly between the treatment and control groups.¹

Physical activity. Whether physical activity interventions slow cognitive decline and prevent dementia in cognitively intact adults was examined in a systematic review of 32 trials.² Most of the trials followed patients for 6 months; a few stretched for 1 or 2 years.

Evidence was insufficient to prove cognitive benefit for short-term, single-component or multicomponent physical activity interventions. However, a multidomain physical activity intervention that also included dietary modifications and cognitive training did show a delay in cognitive decline, but only “low-strength” evidence.²

Nutritional supplements. The antioxidants vitamin E and selenium were studied for their possible cognitive benefit in the double-blind randomized Prevention of Alzheimer Disease by Vitamin E and Selenium trial³ in 3,786 asymptomatic men ages 60 and older. Neither supplement was found to prevent dementia over a 7-year follow-up period.

A review of 38 trials⁴ evaluated the effects on cognition of omega-3 fatty acids, soy, ginkgo biloba, B vitamins, vitamin D plus calcium, vitamin C, beta-carotene, and multi-ingredient supplements. It found insufficient evidence to recommend any over-the-counter supplement for cognitive protection in adults with normal cognition or mild cognitive impairment.

Pharmacologic treatments

Testosterone supplementation. The Testosterone Trials tested the effects of testosterone gel vs placebo for 1 year on 493 men over age 65 with low testosterone (< 275 ng/mL) and with subjective memory complaints and objective memory performance deficits. Treatment was not associated with improved memory or other cognitive functions compared with placebo.⁵

Antiamyloid drugs. A randomized, double-blind, placebo-controlled trial in nearly 2,000 patients evaluated verubecestat, an oral beta-site amyloid precursor protein-cleaving enzyme-1 inhibitor that reduces the amyloid-beta level in cerebrospinal fluid.⁶

Verubecestat did not reduce cognitive or functional decline in patients with mild-to-moderate Alzheimer disease, while adverse events including rashes, falls, injuries, sleep disturbances, suicidal ideation, weight loss, and hair color change were more common in the treatment groups. The trial was terminated early because of futility at 50 months.

And in a placebo-controlled trial of solanezumab, a monoclonal antibody directed against the amyloid beta peptide, no benefit was demonstrated at 80 weeks in more than 2,000 patients with Alzheimer disease.⁷

Multiple common agents. A well-conducted systematic review⁸ of 51 trials of at least a 6-month duration did not support the use of antihypertensive agents, diabetes medications, nonsteroidal anti-inflammatory drugs, aspirin, hormones, or lipid-lowering drugs for cognitive protection for people with normal cognition or mild cognitive impairment.

However, some studies found reassuring evidence that standard therapies for other conditions do not worsen cognitive decline and are protective for atrial fibrillation.⁸

Proton-pump inhibitors. Concern exists for a potential link between dementia risk and proton-pump inhibitors, which are widely used to treat acid-related gastrointestinal disorders.⁹

A prospective population-based cohort study¹⁰ of nearly 3,500 people ages 65 and older without baseline dementia screened participants for dementia every 2 years over a mean period of 7.5 years and provided further evaluation for those who screened positive. Use of proton-pump inhibitors was not found to be associated with dementia risk, even with high cumulative exposure.

Results from this study do not support avoiding proton-pump inhibitors out of concern for dementia risk, although long-term use is associated with other safety concerns.

Oral anticoagulation. The increased risk of dementia with atrial fibrillation is well documented.¹¹

A retrospective study¹² based on a Swedish health registry and using more than 444,000 patients covering more than 1.5 million years at risk found that oral anticoagulant treatment at baseline conferred a 29% lower risk of dementia in an intention-to-treat analysis and a 48% lower risk in on-treatment analysis compared with no oral anticoagulation therapy. No difference was found between new oral anticoagulants and warfarin.

Transcatheter aortic valve implantation is not associated with cognitive decline

For patients with severe aortic stenosis who are not surgical candidates, transcatheter aortic valve implantation is superior to standard medical therapy,¹³ but there are concerns of neurologic and cognitive changes after the procedure.¹⁴ A meta-analysis of 18 studies assessing cognitive performance in more than 1,000 patients (average age ≥ 80) after undergoing the procedure for severe aortic stenosis found no significant cognitive performance changes from baseline perioperatively or 3 or 6 months later.¹⁵

 

 

TREATING DEMENTIA-ASSOCIATED BEHAVIORAL DISTURBANCES

Behavioral and psychiatric symptoms often accompany dementia, but no drugs have yet been approved by the US Food and Drug Administration (FDA) to address them in this population. Nonpharmacologic interventions are recommended as first-line therapy.

Antipsychotics are not recommended

Antipsychotics are often prescribed,¹⁶ although they are associated with metabolic syndrome¹⁷ and increased risks of stroke and death.¹⁸ The FDA has issued black box warnings against using antipsychotics for behavioral management in patients with dementia. Further, the American Geriatrics Society and the American Psychiatric Association do not endorse using them as initial therapy for behavioral and psychological symptoms of dementia.^16,19

The Centers for Medicare and Medicaid Services partnered with nursing homes to improve the quality of care for patients with dementia, with results measured as the rate of prescribing antipsychotic medications. Although the use of psychotropic medications declined after initiating the partnership, the use of mood stabilizers increased, possibly as a substitute for antipsychotics.²⁰

Dextromethorphan-quinidine use is up, despite modest evidence of benefit

A consumer news report in 2017 stated that the use of dextromethorphan-quinidine in long-term care facilities increased by nearly 400% between 2012 and 2016.²¹

Evidence for its benefits comes from a 10-week, phase 2, randomized controlled trial conducted at 42 US study sites with 194 patients with probable Alzheimer disease. Compared with the placebo group, the active treatment group had mildly reduced agitation but an increased risk of falls, dizziness, and diarrhea. However, rates of adverse effects were low, and the authors concluded that treatment was generally well tolerated.²²

Pimavanserin: No long-term benefit for psychosis

In a phase 2, randomized, double-blind, placebo-controlled trial in 181 patients with possible or probable Alzheimer disease and psychotic symptoms, pimavanserin was associated with improved symptoms as measured by the Neuropsychiatric Inventory–Nursing Home Version psychosis score at 6 weeks, but no difference was found compared with placebo at 12 weeks. The treatment group had more adverse events, including agitation, aggression, peripheral edema, anxiety, and symptoms of dementia, although the differences were not statistically significant.²³               

DELIRIUM: AVOID ANTIPSYCHOTICS

Delirium is common in hospitalized older adults, especially those who have baseline cognitive or functional impairment and are exposed to precipitating factors such as treatment with anticholinergic or narcotic medications, infection, surgery, or admission to an intensive care unit.²⁴

Delirium at discharge predicts poor outcomes

In a prospective study of 152 hospitalized patients with delirium, those who either did not recover from delirium or had only partially recovered at discharge were more likely to visit the emergency department, be rehospitalized, or die during the subsequent 3 months than those who had fully recovered from delirium at discharge.²⁵

Multicomponent, patient-centered approach can help

A randomized trial in 377 patients in Taiwan evaluated the use of a modified Hospital Elder Life Program, consisting of 3 protocols focused on orienting communication, oral and nutritional assistance, and early mobilization. Patients were at least 65 years old and undergoing elective abdominal surgery with expected length of hospital stay longer than 6 days. The program, administered daily during hospitalization, significantly lowered postoperative delirium by 56% and hospital stay by 2 days compared with usual care.²⁶

Prophylactic haloperidol does not improve outcomes

In a multicenter randomized, double-blind, placebo-controlled trial, van den Boogaard et al studied prophylactic intravenous haloperidol in nearly 1,800 critically ill patients at high risk of delirium.²⁷ Haloperidol did not improve survival at 28 days compared with placebo. For secondary outcomes, including delirium incidence, delirium-free and coma-free days, duration of mechanical ventilation, and hospital and intensive care department length of stay, treatment was not found to differ statistically from placebo.

Antipsychotics may worsen delirium

A double-blind, parallel-arm, dose-titrated randomized trial, conducted at 11 Australian hospices or hospitals with palliative care services, administered oral risperidone, haloperidol, or placebo to 247 patients with life-limiting illness and delirium. Both treatment groups had higher delirium symptom scores than the placebo group.²⁸

In addition, a systematic review and meta-analysis of 19 studies found no benefit of antipsychotic medications for preventing or treating delirium in hospitalized adults.²⁹

Antipsychotics are often continued indefinitely

A retrospective chart review at a US academic health system found³⁰ that among 487 patients with a new antipsychotic medication prescribed during hospitalization, 147 (30.2%) were discharged on an antipsychotic. Of these, 121 (82.3%) had a diagnosis of delirium. Only 15 (12.4%) had discharge summaries that included instructions for discontinuing the drug.

Another US health system retrospectively reviewed antipsychotic use and found³¹ that out of 260 patients who were newly exposed to an antipsychotic drug during hospitalization, 146 (56.2%) were discharged on an antipsychotic drug, and 65% of these patients were still on the drug at the time of the next hospital admission.

 

 

EXERCISE, EXERCISE, EXERCISE

Exercise recommended, but not vitamin D, to prevent falls

In 2018, the US Preventive Services Task Force updated its recommendations for preventing falls in community-dwelling older adults.³² Based on the findings of several trials, the task force recommends exercise interventions for adults age 65 and older who are at increased risk for falls. Gait, balance, and functional training were studied in 17 trials, resistance training in 13, flexibility in 8, endurance training in 5, and tai chi in 3, with 5 studies including general physical activity. Exercise interventions most commonly took place for 3 sessions per week for 12 months (range 2–42 months).

The task force also recommends against vitamin D supplementation for fall prevention in community-dwelling adults age 65 or older who are not known to have osteoporosis or vitamin D deficiency.

Early mobilization helps inpatients

Hospitalized older adults usually spend most of their time in bed. Forty-five previously ambulatory patients (age ≥ 65 without dementia or delirium) in a Veterans Affairs hospital were monitored with wireless accelerometers and were found to spend, on average, 83% of the measured hospital stay in bed. Standing or walking time ranged from 0.2% to 21%, with a median of only 3% (43 minutes a day).³³

Since falls with injury became a Centers for Medicare and Medicaid Services nonreimbursable hospital-acquired condition, tension has arisen between promoting mobility and preventing falls.³⁴ Two studies evaluating the adoption of mobility-restricting approaches such as bed-alarms, “fall-alert” signs, supervision of patients in the bathroom, and ensuring patients’ walking aids are within reach, did not find a significant reduction in falls or fall-related injuries.^35,36

A clinically significant loss of community mobility is common after hospitalization in older adults.³⁷ Older adults who developed mobility impairment during hospitalization had a higher risk of death in a large, retrospective study.³⁸ A large Canadian multisite intervention trial³⁹ that promoted early mobilization in older patients who were admitted to general medical wards resulted in increased mobilization and significantly shorter hospital stays.

POSTHOSPITAL CARE NEEDS IMPROVEMENT

After hospitalization, older adults who have difficulty with activities of daily living or complex medical needs often require continued care.

About 20% of hospitalized Medicare beneficiaries in the United States are discharged to skilled nursing facilities.⁴⁰ This is often a stressful transition, and most people have little guidance on selecting a facility and simply choose one based on its proximity to home.⁴¹

A program of frequent visits by hospital-employed physicians and advanced practice professionals at skilled nursing facilities resulted in a significantly lower 30-day readmission rate compared with nonparticipating skilled nursing facilities in the same geographic area.⁴²

Home healthcare is recommended after hospital discharge at a rapidly increasing rate. Overall referral rates increased from 8.6% to 14.1% between 2001 and 2012, and from 14.3% to 24.0% for patients with heart failure.⁴³ A qualitative study of home healthcare nurses found a need for improved care coordination between home healthcare agencies and discharging hospitals, including defining accountability for orders and enhancing communication.⁴⁴

Unfortunately, recent research has not unveiled a breakthrough for preventing or treating cognitive impairment or Alzheimer disease. But several studies from the last 2 years are helping to drive the field of geriatrics forward, providing evidence of what does and does not help a variety of issues specific to the elderly. 

Based on a search of the 2017 and 2018 literature, this article presents new evidence on preventing and treating cognitive impairment, managing dementia-associated behavioral disturbances and delirium, preventing falls, and improving inpatient mobility and posthospital care transitions.

COGNITIVE IMPAIRMENT, DEMENTIA: STILL NO SILVER BULLET

With the exception of oral anticoagulation treatment for atrial fibrillation, there is little evidence that pharmacologic or nonpharmacologic interventions slow the onset or progression of Alzheimer disease.

Nonpharmacologic interventions

Home occupational therapy. A 2-year home-based occupational therapy intervention¹ showed no evidence of slowing functional decline in patients with Alzheimer disease. The randomized controlled trial involving 180 participants consisted of monthly sessions of an intensive, well-established collaborative-care management model that included fall prevention and other safety strategies, personalized training in activities of daily living, exercise, and education. Outcome measures for activities of daily living did not differ significantly between the treatment and control groups.¹

Physical activity. Whether physical activity interventions slow cognitive decline and prevent dementia in cognitively intact adults was examined in a systematic review of 32 trials.² Most of the trials followed patients for 6 months; a few stretched for 1 or 2 years.

Evidence was insufficient to prove cognitive benefit for short-term, single-component or multicomponent physical activity interventions. However, a multidomain physical activity intervention that also included dietary modifications and cognitive training did show a delay in cognitive decline, but only “low-strength” evidence.²

Nutritional supplements. The antioxidants vitamin E and selenium were studied for their possible cognitive benefit in the double-blind randomized Prevention of Alzheimer Disease by Vitamin E and Selenium trial³ in 3,786 asymptomatic men ages 60 and older. Neither supplement was found to prevent dementia over a 7-year follow-up period.

A review of 38 trials⁴ evaluated the effects on cognition of omega-3 fatty acids, soy, ginkgo biloba, B vitamins, vitamin D plus calcium, vitamin C, beta-carotene, and multi-ingredient supplements. It found insufficient evidence to recommend any over-the-counter supplement for cognitive protection in adults with normal cognition or mild cognitive impairment.

Pharmacologic treatments

Testosterone supplementation. The Testosterone Trials tested the effects of testosterone gel vs placebo for 1 year on 493 men over age 65 with low testosterone (< 275 ng/mL) and with subjective memory complaints and objective memory performance deficits. Treatment was not associated with improved memory or other cognitive functions compared with placebo.⁵

Antiamyloid drugs. A randomized, double-blind, placebo-controlled trial in nearly 2,000 patients evaluated verubecestat, an oral beta-site amyloid precursor protein-cleaving enzyme-1 inhibitor that reduces the amyloid-beta level in cerebrospinal fluid.⁶

Verubecestat did not reduce cognitive or functional decline in patients with mild-to-moderate Alzheimer disease, while adverse events including rashes, falls, injuries, sleep disturbances, suicidal ideation, weight loss, and hair color change were more common in the treatment groups. The trial was terminated early because of futility at 50 months.

And in a placebo-controlled trial of solanezumab, a monoclonal antibody directed against the amyloid beta peptide, no benefit was demonstrated at 80 weeks in more than 2,000 patients with Alzheimer disease.⁷

Multiple common agents. A well-conducted systematic review⁸ of 51 trials of at least a 6-month duration did not support the use of antihypertensive agents, diabetes medications, nonsteroidal anti-inflammatory drugs, aspirin, hormones, or lipid-lowering drugs for cognitive protection for people with normal cognition or mild cognitive impairment.

However, some studies found reassuring evidence that standard therapies for other conditions do not worsen cognitive decline and are protective for atrial fibrillation.⁸

Proton-pump inhibitors. Concern exists for a potential link between dementia risk and proton-pump inhibitors, which are widely used to treat acid-related gastrointestinal disorders.⁹

A prospective population-based cohort study¹⁰ of nearly 3,500 people ages 65 and older without baseline dementia screened participants for dementia every 2 years over a mean period of 7.5 years and provided further evaluation for those who screened positive. Use of proton-pump inhibitors was not found to be associated with dementia risk, even with high cumulative exposure.

Results from this study do not support avoiding proton-pump inhibitors out of concern for dementia risk, although long-term use is associated with other safety concerns.

Oral anticoagulation. The increased risk of dementia with atrial fibrillation is well documented.¹¹

A retrospective study¹² based on a Swedish health registry and using more than 444,000 patients covering more than 1.5 million years at risk found that oral anticoagulant treatment at baseline conferred a 29% lower risk of dementia in an intention-to-treat analysis and a 48% lower risk in on-treatment analysis compared with no oral anticoagulation therapy. No difference was found between new oral anticoagulants and warfarin.

Transcatheter aortic valve implantation is not associated with cognitive decline

For patients with severe aortic stenosis who are not surgical candidates, transcatheter aortic valve implantation is superior to standard medical therapy,¹³ but there are concerns of neurologic and cognitive changes after the procedure.¹⁴ A meta-analysis of 18 studies assessing cognitive performance in more than 1,000 patients (average age ≥ 80) after undergoing the procedure for severe aortic stenosis found no significant cognitive performance changes from baseline perioperatively or 3 or 6 months later.¹⁵

 

 

TREATING DEMENTIA-ASSOCIATED BEHAVIORAL DISTURBANCES

Behavioral and psychiatric symptoms often accompany dementia, but no drugs have yet been approved by the US Food and Drug Administration (FDA) to address them in this population. Nonpharmacologic interventions are recommended as first-line therapy.

Antipsychotics are not recommended

Antipsychotics are often prescribed,¹⁶ although they are associated with metabolic syndrome¹⁷ and increased risks of stroke and death.¹⁸ The FDA has issued black box warnings against using antipsychotics for behavioral management in patients with dementia. Further, the American Geriatrics Society and the American Psychiatric Association do not endorse using them as initial therapy for behavioral and psychological symptoms of dementia.^16,19

The Centers for Medicare and Medicaid Services partnered with nursing homes to improve the quality of care for patients with dementia, with results measured as the rate of prescribing antipsychotic medications. Although the use of psychotropic medications declined after initiating the partnership, the use of mood stabilizers increased, possibly as a substitute for antipsychotics.²⁰

Dextromethorphan-quinidine use is up, despite modest evidence of benefit

A consumer news report in 2017 stated that the use of dextromethorphan-quinidine in long-term care facilities increased by nearly 400% between 2012 and 2016.²¹

Evidence for its benefits comes from a 10-week, phase 2, randomized controlled trial conducted at 42 US study sites with 194 patients with probable Alzheimer disease. Compared with the placebo group, the active treatment group had mildly reduced agitation but an increased risk of falls, dizziness, and diarrhea. However, rates of adverse effects were low, and the authors concluded that treatment was generally well tolerated.²²

Pimavanserin: No long-term benefit for psychosis

In a phase 2, randomized, double-blind, placebo-controlled trial in 181 patients with possible or probable Alzheimer disease and psychotic symptoms, pimavanserin was associated with improved symptoms as measured by the Neuropsychiatric Inventory–Nursing Home Version psychosis score at 6 weeks, but no difference was found compared with placebo at 12 weeks. The treatment group had more adverse events, including agitation, aggression, peripheral edema, anxiety, and symptoms of dementia, although the differences were not statistically significant.²³               

DELIRIUM: AVOID ANTIPSYCHOTICS

Delirium is common in hospitalized older adults, especially those who have baseline cognitive or functional impairment and are exposed to precipitating factors such as treatment with anticholinergic or narcotic medications, infection, surgery, or admission to an intensive care unit.²⁴

Delirium at discharge predicts poor outcomes

In a prospective study of 152 hospitalized patients with delirium, those who either did not recover from delirium or had only partially recovered at discharge were more likely to visit the emergency department, be rehospitalized, or die during the subsequent 3 months than those who had fully recovered from delirium at discharge.²⁵

Multicomponent, patient-centered approach can help

A randomized trial in 377 patients in Taiwan evaluated the use of a modified Hospital Elder Life Program, consisting of 3 protocols focused on orienting communication, oral and nutritional assistance, and early mobilization. Patients were at least 65 years old and undergoing elective abdominal surgery with expected length of hospital stay longer than 6 days. The program, administered daily during hospitalization, significantly lowered postoperative delirium by 56% and hospital stay by 2 days compared with usual care.²⁶

Prophylactic haloperidol does not improve outcomes

In a multicenter randomized, double-blind, placebo-controlled trial, van den Boogaard et al studied prophylactic intravenous haloperidol in nearly 1,800 critically ill patients at high risk of delirium.²⁷ Haloperidol did not improve survival at 28 days compared with placebo. For secondary outcomes, including delirium incidence, delirium-free and coma-free days, duration of mechanical ventilation, and hospital and intensive care department length of stay, treatment was not found to differ statistically from placebo.

Antipsychotics may worsen delirium

A double-blind, parallel-arm, dose-titrated randomized trial, conducted at 11 Australian hospices or hospitals with palliative care services, administered oral risperidone, haloperidol, or placebo to 247 patients with life-limiting illness and delirium. Both treatment groups had higher delirium symptom scores than the placebo group.²⁸

In addition, a systematic review and meta-analysis of 19 studies found no benefit of antipsychotic medications for preventing or treating delirium in hospitalized adults.²⁹

Antipsychotics are often continued indefinitely

A retrospective chart review at a US academic health system found³⁰ that among 487 patients with a new antipsychotic medication prescribed during hospitalization, 147 (30.2%) were discharged on an antipsychotic. Of these, 121 (82.3%) had a diagnosis of delirium. Only 15 (12.4%) had discharge summaries that included instructions for discontinuing the drug.

Another US health system retrospectively reviewed antipsychotic use and found³¹ that out of 260 patients who were newly exposed to an antipsychotic drug during hospitalization, 146 (56.2%) were discharged on an antipsychotic drug, and 65% of these patients were still on the drug at the time of the next hospital admission.

 

 

EXERCISE, EXERCISE, EXERCISE

Exercise recommended, but not vitamin D, to prevent falls

In 2018, the US Preventive Services Task Force updated its recommendations for preventing falls in community-dwelling older adults.³² Based on the findings of several trials, the task force recommends exercise interventions for adults age 65 and older who are at increased risk for falls. Gait, balance, and functional training were studied in 17 trials, resistance training in 13, flexibility in 8, endurance training in 5, and tai chi in 3, with 5 studies including general physical activity. Exercise interventions most commonly took place for 3 sessions per week for 12 months (range 2–42 months).

The task force also recommends against vitamin D supplementation for fall prevention in community-dwelling adults age 65 or older who are not known to have osteoporosis or vitamin D deficiency.

Early mobilization helps inpatients

Hospitalized older adults usually spend most of their time in bed. Forty-five previously ambulatory patients (age ≥ 65 without dementia or delirium) in a Veterans Affairs hospital were monitored with wireless accelerometers and were found to spend, on average, 83% of the measured hospital stay in bed. Standing or walking time ranged from 0.2% to 21%, with a median of only 3% (43 minutes a day).³³

Since falls with injury became a Centers for Medicare and Medicaid Services nonreimbursable hospital-acquired condition, tension has arisen between promoting mobility and preventing falls.³⁴ Two studies evaluating the adoption of mobility-restricting approaches such as bed-alarms, “fall-alert” signs, supervision of patients in the bathroom, and ensuring patients’ walking aids are within reach, did not find a significant reduction in falls or fall-related injuries.^35,36

A clinically significant loss of community mobility is common after hospitalization in older adults.³⁷ Older adults who developed mobility impairment during hospitalization had a higher risk of death in a large, retrospective study.³⁸ A large Canadian multisite intervention trial³⁹ that promoted early mobilization in older patients who were admitted to general medical wards resulted in increased mobilization and significantly shorter hospital stays.

POSTHOSPITAL CARE NEEDS IMPROVEMENT

After hospitalization, older adults who have difficulty with activities of daily living or complex medical needs often require continued care.

About 20% of hospitalized Medicare beneficiaries in the United States are discharged to skilled nursing facilities.⁴⁰ This is often a stressful transition, and most people have little guidance on selecting a facility and simply choose one based on its proximity to home.⁴¹

A program of frequent visits by hospital-employed physicians and advanced practice professionals at skilled nursing facilities resulted in a significantly lower 30-day readmission rate compared with nonparticipating skilled nursing facilities in the same geographic area.⁴²

Home healthcare is recommended after hospital discharge at a rapidly increasing rate. Overall referral rates increased from 8.6% to 14.1% between 2001 and 2012, and from 14.3% to 24.0% for patients with heart failure.⁴³ A qualitative study of home healthcare nurses found a need for improved care coordination between home healthcare agencies and discharging hospitals, including defining accountability for orders and enhancing communication.⁴⁴

Unfortunately, recent research has not unveiled a breakthrough for preventing or treating cognitive impairment or Alzheimer disease. But several studies from the last 2 years are helping to drive the field of geriatrics forward, providing evidence of what does and does not help a variety of issues specific to the elderly. 

Based on a search of the 2017 and 2018 literature, this article presents new evidence on preventing and treating cognitive impairment, managing dementia-associated behavioral disturbances and delirium, preventing falls, and improving inpatient mobility and posthospital care transitions.

COGNITIVE IMPAIRMENT, DEMENTIA: STILL NO SILVER BULLET

With the exception of oral anticoagulation treatment for atrial fibrillation, there is little evidence that pharmacologic or nonpharmacologic interventions slow the onset or progression of Alzheimer disease.

Nonpharmacologic interventions

Home occupational therapy. A 2-year home-based occupational therapy intervention¹ showed no evidence of slowing functional decline in patients with Alzheimer disease. The randomized controlled trial involving 180 participants consisted of monthly sessions of an intensive, well-established collaborative-care management model that included fall prevention and other safety strategies, personalized training in activities of daily living, exercise, and education. Outcome measures for activities of daily living did not differ significantly between the treatment and control groups.¹

Physical activity. Whether physical activity interventions slow cognitive decline and prevent dementia in cognitively intact adults was examined in a systematic review of 32 trials.² Most of the trials followed patients for 6 months; a few stretched for 1 or 2 years.

Evidence was insufficient to prove cognitive benefit for short-term, single-component or multicomponent physical activity interventions. However, a multidomain physical activity intervention that also included dietary modifications and cognitive training did show a delay in cognitive decline, but only “low-strength” evidence.²

Nutritional supplements. The antioxidants vitamin E and selenium were studied for their possible cognitive benefit in the double-blind randomized Prevention of Alzheimer Disease by Vitamin E and Selenium trial³ in 3,786 asymptomatic men ages 60 and older. Neither supplement was found to prevent dementia over a 7-year follow-up period.

A review of 38 trials⁴ evaluated the effects on cognition of omega-3 fatty acids, soy, ginkgo biloba, B vitamins, vitamin D plus calcium, vitamin C, beta-carotene, and multi-ingredient supplements. It found insufficient evidence to recommend any over-the-counter supplement for cognitive protection in adults with normal cognition or mild cognitive impairment.

Pharmacologic treatments

Testosterone supplementation. The Testosterone Trials tested the effects of testosterone gel vs placebo for 1 year on 493 men over age 65 with low testosterone (< 275 ng/mL) and with subjective memory complaints and objective memory performance deficits. Treatment was not associated with improved memory or other cognitive functions compared with placebo.⁵

Antiamyloid drugs. A randomized, double-blind, placebo-controlled trial in nearly 2,000 patients evaluated verubecestat, an oral beta-site amyloid precursor protein-cleaving enzyme-1 inhibitor that reduces the amyloid-beta level in cerebrospinal fluid.⁶

Verubecestat did not reduce cognitive or functional decline in patients with mild-to-moderate Alzheimer disease, while adverse events including rashes, falls, injuries, sleep disturbances, suicidal ideation, weight loss, and hair color change were more common in the treatment groups. The trial was terminated early because of futility at 50 months.

And in a placebo-controlled trial of solanezumab, a monoclonal antibody directed against the amyloid beta peptide, no benefit was demonstrated at 80 weeks in more than 2,000 patients with Alzheimer disease.⁷

Multiple common agents. A well-conducted systematic review⁸ of 51 trials of at least a 6-month duration did not support the use of antihypertensive agents, diabetes medications, nonsteroidal anti-inflammatory drugs, aspirin, hormones, or lipid-lowering drugs for cognitive protection for people with normal cognition or mild cognitive impairment.

However, some studies found reassuring evidence that standard therapies for other conditions do not worsen cognitive decline and are protective for atrial fibrillation.⁸

Proton-pump inhibitors. Concern exists for a potential link between dementia risk and proton-pump inhibitors, which are widely used to treat acid-related gastrointestinal disorders.⁹

A prospective population-based cohort study¹⁰ of nearly 3,500 people ages 65 and older without baseline dementia screened participants for dementia every 2 years over a mean period of 7.5 years and provided further evaluation for those who screened positive. Use of proton-pump inhibitors was not found to be associated with dementia risk, even with high cumulative exposure.

Results from this study do not support avoiding proton-pump inhibitors out of concern for dementia risk, although long-term use is associated with other safety concerns.

Oral anticoagulation. The increased risk of dementia with atrial fibrillation is well documented.¹¹

A retrospective study¹² based on a Swedish health registry and using more than 444,000 patients covering more than 1.5 million years at risk found that oral anticoagulant treatment at baseline conferred a 29% lower risk of dementia in an intention-to-treat analysis and a 48% lower risk in on-treatment analysis compared with no oral anticoagulation therapy. No difference was found between new oral anticoagulants and warfarin.

Transcatheter aortic valve implantation is not associated with cognitive decline

For patients with severe aortic stenosis who are not surgical candidates, transcatheter aortic valve implantation is superior to standard medical therapy,¹³ but there are concerns of neurologic and cognitive changes after the procedure.¹⁴ A meta-analysis of 18 studies assessing cognitive performance in more than 1,000 patients (average age ≥ 80) after undergoing the procedure for severe aortic stenosis found no significant cognitive performance changes from baseline perioperatively or 3 or 6 months later.¹⁵

 

 

TREATING DEMENTIA-ASSOCIATED BEHAVIORAL DISTURBANCES

Behavioral and psychiatric symptoms often accompany dementia, but no drugs have yet been approved by the US Food and Drug Administration (FDA) to address them in this population. Nonpharmacologic interventions are recommended as first-line therapy.

Antipsychotics are not recommended

Antipsychotics are often prescribed,¹⁶ although they are associated with metabolic syndrome¹⁷ and increased risks of stroke and death.¹⁸ The FDA has issued black box warnings against using antipsychotics for behavioral management in patients with dementia. Further, the American Geriatrics Society and the American Psychiatric Association do not endorse using them as initial therapy for behavioral and psychological symptoms of dementia.^16,19

The Centers for Medicare and Medicaid Services partnered with nursing homes to improve the quality of care for patients with dementia, with results measured as the rate of prescribing antipsychotic medications. Although the use of psychotropic medications declined after initiating the partnership, the use of mood stabilizers increased, possibly as a substitute for antipsychotics.²⁰

Dextromethorphan-quinidine use is up, despite modest evidence of benefit

A consumer news report in 2017 stated that the use of dextromethorphan-quinidine in long-term care facilities increased by nearly 400% between 2012 and 2016.²¹

Evidence for its benefits comes from a 10-week, phase 2, randomized controlled trial conducted at 42 US study sites with 194 patients with probable Alzheimer disease. Compared with the placebo group, the active treatment group had mildly reduced agitation but an increased risk of falls, dizziness, and diarrhea. However, rates of adverse effects were low, and the authors concluded that treatment was generally well tolerated.²²

Pimavanserin: No long-term benefit for psychosis

In a phase 2, randomized, double-blind, placebo-controlled trial in 181 patients with possible or probable Alzheimer disease and psychotic symptoms, pimavanserin was associated with improved symptoms as measured by the Neuropsychiatric Inventory–Nursing Home Version psychosis score at 6 weeks, but no difference was found compared with placebo at 12 weeks. The treatment group had more adverse events, including agitation, aggression, peripheral edema, anxiety, and symptoms of dementia, although the differences were not statistically significant.²³               

DELIRIUM: AVOID ANTIPSYCHOTICS

Delirium is common in hospitalized older adults, especially those who have baseline cognitive or functional impairment and are exposed to precipitating factors such as treatment with anticholinergic or narcotic medications, infection, surgery, or admission to an intensive care unit.²⁴

Delirium at discharge predicts poor outcomes

In a prospective study of 152 hospitalized patients with delirium, those who either did not recover from delirium or had only partially recovered at discharge were more likely to visit the emergency department, be rehospitalized, or die during the subsequent 3 months than those who had fully recovered from delirium at discharge.²⁵

Multicomponent, patient-centered approach can help

A randomized trial in 377 patients in Taiwan evaluated the use of a modified Hospital Elder Life Program, consisting of 3 protocols focused on orienting communication, oral and nutritional assistance, and early mobilization. Patients were at least 65 years old and undergoing elective abdominal surgery with expected length of hospital stay longer than 6 days. The program, administered daily during hospitalization, significantly lowered postoperative delirium by 56% and hospital stay by 2 days compared with usual care.²⁶

Prophylactic haloperidol does not improve outcomes

In a multicenter randomized, double-blind, placebo-controlled trial, van den Boogaard et al studied prophylactic intravenous haloperidol in nearly 1,800 critically ill patients at high risk of delirium.²⁷ Haloperidol did not improve survival at 28 days compared with placebo. For secondary outcomes, including delirium incidence, delirium-free and coma-free days, duration of mechanical ventilation, and hospital and intensive care department length of stay, treatment was not found to differ statistically from placebo.

Antipsychotics may worsen delirium

A double-blind, parallel-arm, dose-titrated randomized trial, conducted at 11 Australian hospices or hospitals with palliative care services, administered oral risperidone, haloperidol, or placebo to 247 patients with life-limiting illness and delirium. Both treatment groups had higher delirium symptom scores than the placebo group.²⁸

In addition, a systematic review and meta-analysis of 19 studies found no benefit of antipsychotic medications for preventing or treating delirium in hospitalized adults.²⁹

Antipsychotics are often continued indefinitely

A retrospective chart review at a US academic health system found³⁰ that among 487 patients with a new antipsychotic medication prescribed during hospitalization, 147 (30.2%) were discharged on an antipsychotic. Of these, 121 (82.3%) had a diagnosis of delirium. Only 15 (12.4%) had discharge summaries that included instructions for discontinuing the drug.

Another US health system retrospectively reviewed antipsychotic use and found³¹ that out of 260 patients who were newly exposed to an antipsychotic drug during hospitalization, 146 (56.2%) were discharged on an antipsychotic drug, and 65% of these patients were still on the drug at the time of the next hospital admission.

 

 

EXERCISE, EXERCISE, EXERCISE

Exercise recommended, but not vitamin D, to prevent falls

In 2018, the US Preventive Services Task Force updated its recommendations for preventing falls in community-dwelling older adults.³² Based on the findings of several trials, the task force recommends exercise interventions for adults age 65 and older who are at increased risk for falls. Gait, balance, and functional training were studied in 17 trials, resistance training in 13, flexibility in 8, endurance training in 5, and tai chi in 3, with 5 studies including general physical activity. Exercise interventions most commonly took place for 3 sessions per week for 12 months (range 2–42 months).

The task force also recommends against vitamin D supplementation for fall prevention in community-dwelling adults age 65 or older who are not known to have osteoporosis or vitamin D deficiency.

Early mobilization helps inpatients

Hospitalized older adults usually spend most of their time in bed. Forty-five previously ambulatory patients (age ≥ 65 without dementia or delirium) in a Veterans Affairs hospital were monitored with wireless accelerometers and were found to spend, on average, 83% of the measured hospital stay in bed. Standing or walking time ranged from 0.2% to 21%, with a median of only 3% (43 minutes a day).³³

Since falls with injury became a Centers for Medicare and Medicaid Services nonreimbursable hospital-acquired condition, tension has arisen between promoting mobility and preventing falls.³⁴ Two studies evaluating the adoption of mobility-restricting approaches such as bed-alarms, “fall-alert” signs, supervision of patients in the bathroom, and ensuring patients’ walking aids are within reach, did not find a significant reduction in falls or fall-related injuries.^35,36

A clinically significant loss of community mobility is common after hospitalization in older adults.³⁷ Older adults who developed mobility impairment during hospitalization had a higher risk of death in a large, retrospective study.³⁸ A large Canadian multisite intervention trial³⁹ that promoted early mobilization in older patients who were admitted to general medical wards resulted in increased mobilization and significantly shorter hospital stays.

POSTHOSPITAL CARE NEEDS IMPROVEMENT

After hospitalization, older adults who have difficulty with activities of daily living or complex medical needs often require continued care.

About 20% of hospitalized Medicare beneficiaries in the United States are discharged to skilled nursing facilities.⁴⁰ This is often a stressful transition, and most people have little guidance on selecting a facility and simply choose one based on its proximity to home.⁴¹

A program of frequent visits by hospital-employed physicians and advanced practice professionals at skilled nursing facilities resulted in a significantly lower 30-day readmission rate compared with nonparticipating skilled nursing facilities in the same geographic area.⁴²

Home healthcare is recommended after hospital discharge at a rapidly increasing rate. Overall referral rates increased from 8.6% to 14.1% between 2001 and 2012, and from 14.3% to 24.0% for patients with heart failure.⁴³ A qualitative study of home healthcare nurses found a need for improved care coordination between home healthcare agencies and discharging hospitals, including defining accountability for orders and enhancing communication.⁴⁴

Narcolepsy was originally described in the late 1800s by the French physician Jean-Baptiste-Edouard Gélineau, who reported the case of a wine merchant suffering from somnolence. In this first description, he coined the term narcolepsie by joining the Greek words narke (numbness or stupor) and lepsis (attack).¹

Since then, the disorder has been further characterized, and some insight into its biological underpinnings has been established. Importantly, treatments have improved and expanded, facilitating its management and thereby improving quality of life for those with the disorder.

This review focuses on clinically relevant features of the disorder and proposes management strategies.

CLINICAL FEATURES

Narcolepsy is characterized by instability of sleep-wake transitions.

Daytime sleepiness

Clinically, narcolepsy manifests with excessive daytime sleepiness that can be personally and socially disabling. Cataplexy, sleep paralysis, and hypnagogic or hypnopompic hallucinations can also be present,^2,3 but they are not necessary for diagnosis. In fact, a minority of patients with narcolepsy have all these symptoms.⁴ Narcolepsy is divided into type 1 (with cataplexy) and type 2 (without cataplexy).²

Sleepiness tends to be worse with inactivity, and sleep can often be irresistible. Sleep attacks can come on suddenly and may be brief enough to manifest as a lapse in consciousness.

Short naps tend to be refreshing. Rapid eye movement (REM) latency—the interval between falling asleep and the onset of the REM sleep—is short in narcolepsy, and since the REM stage is when dreaming occurs, naps often include dreaming. Therefore, when taking a history, it is worthwhile to ask patients whether they dream during naps; a yes answer supports the diagnosis of narcolepsy.⁵

In children, sleepiness can manifest in reduced concentration and behavioral issues.⁶ Napping after age 5 or 6 is considered abnormal and may reflect pathologic sleepiness.⁷

Cataplexy

Cataplexy—transient muscle weakness triggered by emotion—is a specific feature of narcolepsy type 1. It often begins in the facial muscles and can manifest with slackening of the jaw or brief dropping of the head. However, episodes can be more dramatic and, if the trunk and limb muscles are affected, can result in collapsing to the ground.

Cataplexy usually has its onset at about the same time as the sleepiness associated with narcolepsy, but it can arise even years later.⁸ Episodes can last from a few seconds to 2 minutes. Consciousness is always preserved. A range of emotions can trigger cataplexy, but typically the emotion is a positive one such as laughter or excitement.⁹ Deep tendon reflexes disappear in cataplexy, so checking reflexes during a witnessed episode can be clinically valuable.²

Cataplexy can worsen with stress and insufficient sleep, occasionally with “status cataplecticus,” in which repeated, persistent episodes of cataplexy occur over several hours.⁸ Status cataplecticus can be spontaneous or an effect of withdrawal from anticataplectic medications.²

Cataplexy is thought to represent intrusion of REM sleep and its associated muscle atonia during wakefulness.

Sleep paralysis, hallucinations

Sleep paralysis and hallucinations are other features of narcolepsy that reflect this REM dissociation from sleep.

Sleep paralysis occurs most commonly upon awakening, but sometimes just before sleep onset. In most cases, it is manifested by inability to move the limbs or speak, lasting several seconds or, in rare cases, minutes at a time. Sleep paralysis can be associated with a sensation of fear or suffocation, especially when initially experienced.⁸

Hypnopompic hallucinations, occurring upon awakening, are more common than hypnagogic hallucinations, which are experienced before falling asleep. The hallucinations are often vivid and usually visual, although other types of hallucinations are possible. Unlike those that occur in psychotic disorders, the hallucinations tend to be associated with preserved insight that they are not real.¹⁰

Notably, both sleep paralysis and hallucinations are nonspecific symptoms that are common in the general population.^8,11,12

Fragmented sleep

Although they are very sleepy, people with narcolepsy generally cannot stay asleep for very long. Their sleep tends to be extremely fragmented, and they often wake up several times a night.²

This sleep pattern reflects the inherent instability of sleep-wake transitions in narcolepsy. In fact, over a 24-hour period, adults with narcolepsy have a normal amount of sleep.¹³ In children, however, when narcolepsy first arises, the 24-hour sleep time can increase abruptly and can sometimes be associated with persistent cataplexy that can manifest as a clumsy gait.¹⁴

Weight gain, obstructive sleep apnea

Weight gain is common, particularly after symptom onset, and especially in children. As a result, obesity is a frequent comorbidity.¹⁵ Because obstructive sleep apnea can consequently develop, all patients with narcolepsy require screening for sleep-disordered breathing.

Other sleep disorders often accompany narcolepsy and are more common than in the general population.¹⁶ In a study incorporating both clinical and polysomnographic data of 100 patients with narcolepsy, insomnia was the most common comorbid sleep disorder, with a prevalence of 28%; others were REM sleep behavior disorder (24%), restless legs syndrome (24%), obstructive sleep apnea (21%), and non-REM parasomnias.¹⁷

 

 

PSYCHOSOCIAL CONSEQUENCES

Narcolepsy has significant psychosocial consequences. As a result of their symptoms, people with narcolepsy may not be able to meet academic or work-related demands.

Additionally, their risk of a motor vehicle accident is 3 to 4 times higher than in the general population, and more than one-third of patients have been in an accident due to sleepiness.¹⁸ There is some evidence to show that treatment eliminates this risk.¹⁹

Few systematic studies have examined mood disorders in narcolepsy. However, studies tend to show a higher prevalence of psychiatric disorders than in the general population, with depression and anxiety the most com-mon.^20,21

DIAGNOSIS IS OFTEN DELAYED

The prevalence of narcolepsy type 1 is between 25 and 100 per 100,000 people.²² In a Mayo Clinic study,²³ the incidence of narcolepsy type 1 was estimated to be 0.74 per 100,000 person-years. Epidemiologic data on narcolepsy type 2 are sparse, but patients with narcolepsy without cataplexy are thought to represent only 36% of all narcolepsy patients.²³

Diagnosis is often delayed, with the average time between the onset of symptoms and the diagnosis ranging from 8 to 22 years. With increasing awareness, the efficiency of the diagnostic process is improving, and this delay is expected to lessen accordingly.²⁴

Symptoms most commonly arise in the second decade; but the age at onset ranges significantly, between the first and fifth decades. Narcolepsy has a bimodal distribution in incidence, with the biggest peak at approximately age 15 and second smaller peak in the mid-30s. Some studies have suggested a slight male predominance.^23,25

DIAGNOSIS

History is key

The history should include specific questions about the hallmark features of narcolepsy, including cataplexy, sleep paralysis, and sleep-related hallucinations. For individual assessment of subjective sleepiness, the Epworth Sleepiness Scale or Pediatric Daytime Sleepiness Scale can be administered quickly in the office setting.^26,27

The Epworth score is calculated from the self-rated likelihood of falling asleep in 8 different situations, with possible scores of 0 (would never doze) to 3 (high chance of dozing) on each question, for a total possible score of 0 to 24. Normal total scores are between 0 and 10, while scores greater than 10 reflect pathologic sleepiness. Scores on the Epworth Sleepiness Scale in those with narcolepsy tend to reflect moderate to severe sleepiness, or at least 13, as opposed to patients with obstructive sleep apnea, whose scores commonly reflect milder sleepiness.²⁸

Testing with actigraphy and polysomnography

It is imperative to rule out insufficient sleep and other sleep disorders as a cause of daytime sleepiness. This can be done with a careful clinical history, actigraphy with sleep logs, and polysomnography.

In the 2 to 4 weeks before actigraphy and subsequent testing, all medications with alerting or sedating properties (including anti­depressants) should be tapered off to prevent influence on the results of the study.

Delayed sleep-phase disorder presents at a similar age as narcolepsy and can be associated with similar degrees of sleepiness. However, individuals with delayed sleep phase disorder have an inappropriately timed sleep-wake cycle so that there is a shift in their desired sleep onset and awakening times. It is common—prevalence estimates vary but average about 1% in the general population.²⁹

Insufficient sleep syndrome is even more common, especially in teenagers and young adults, with increasing family, social, and academic demands. Sleep needs vary across the life span. A teenager needs 8 to 10 hours of sleep per night, and a young adult needs 7 to 9 hours. A study of 1,285 high school students found that 10.4% were not getting enough sleep.³⁰

If actigraphy data suggest a circadian rhythm disorder or insufficient sleep that could explain the symptoms of sleepiness, then further testing should be halted and these specific issues should be addressed. In these cases, working with the patient toward maintaining a regular sleep-wake schedule with 7 to 8 hours of nightly sleep will often resolve symptoms.

If actigraphy demonstrates the patient is maintaining a regular sleep schedule and allowing adequate time for nightly sleep, the next step is polysomnography.

Polysomnography is performed to detect other disorders that can disrupt sleep, such as sleep-disordered breathing or periodic limb movement disorder.^2,5 In addition, polysomnography can provide assurance that adequate sleep was obtained prior to the next step in testing.

Multiple sleep latency test

If sufficient sleep is obtained on polysomnograpy (at least 6 hours for an adult) and no other sleep disorder is identified, a multiple sleep latency test is performed. A urine toxicology screen is typically performed on the day of the test to ensure that drugs are not affecting the results.

The multiple sleep latency test consists of 4 to 5 nap opportunities at 2-hour intervals in a quiet dark room conducive to sleep, during which both sleep and REM latency are recorded. The sleep latency of those with narcolepsy is significantly shortened, and the diagnosis of narcolepsy requires an average sleep latency of less than 8 minutes.

Given the propensity for REM sleep in narcolepsy, another essential feature for diagnosis is the sleep-onset REM period (SOREMP). A SOREMP is defined as a REM latency of less than 15 minutes. A diagnosis of narcolepsy re-quires a SOREMP in at least 2 of the naps in a multiple sleep latency test (or 1 nap if the shortened REM latency is seen during polysomnography).³¹

The multiple sleep latency test has an imperfect sensitivity, though, and should be repeated when there is a high suspicion of narcolepsy.^32–34 It is not completely specific either, and false-positive results occur. In fact, SOREMPs can be seen in the general population, particularly in those with a circadian rhythm disorder, insufficient sleep, or sleep-disordered breathing. Two or more SOREMPs in an multiple sleep latency test can be seen in a small proportion of the general population.³⁵ The results of a multiple sleep latency test should be interpreted in the clinical context.

Differential diagnosis

Narcolepsy type 1 is distinguished from type 2 by the presence of cataplexy. A cerebrospinal fluid hypocretin 1 level of 110 pg/mL or less, or less than one-third of the mean value obtained in normal individuals, can substitute for the multiple sleep latency test in diagnosing narcolepsy type 1.³¹ Currently, hypocretin testing is generally not performed in clinical practice, although it may become a routine part of the narcolepsy evaluation in the future.

Thus, according to the International Classification of Sleep Disorders, 3rd edition,³¹ the diagnosis of narcolepsy type 1 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurologic disorder, mental disorder, medication use, or substance use disorder, and at least 1 of the following:

• Cataplexy and mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing (1 of which can be on the preceding night’s polysomography)
• Cerebrospinal fluid hypocretin 1 levels less than 110 pg/mL or one-third the baseline normal levels and mean sleep latency ≤ 8 minutes with ≥ 2 SOREMPs on multiple sleep latency testing.

Similarly, the diagnosis of narcolepsy type 2 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurological disorder, mental disorder, medication use, or substance use disorder, plus:

• Mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing.

Idiopathic hypersomnia, another disorder of central hypersomnolence, is also characterized by disabling sleepiness. It is diagnostically differentiated from narcolepsy, as there are fewer than 2 SOREMPs. As opposed to narcolepsy, in which naps tend to be refreshing, even prolonged naps in idiopathic hypersomnia are often not helpful in restoring wakefulness. In idiopathic hypersomnia, sleep is usually not fragmented, and there are few nocturnal arousals. Sleep times can often be prolonged as well, whereas in narcolepsy total sleep time through the day may not be increased but is not consolidated.

Kleine-Levin syndrome is a rarer disorder of hypersomnia. It is episodic compared with the relatively persistent sleepiness in narcolepsy and idiopathic hypersomnia. Periods of hypersomnia occur intermittently for days to weeks and are accompanied by cognitive and behavioral changes including hyperphagia and hypersexuality.⁴

LINKED TO HYPOCRETIN DEFICIENCY

Over the past 2 decades, the underlying pathophysiology of narcolepsy type 1 has been better characterized.

Narcolepsy type 1 has been linked to a deficiency in hypocretin in the central nervous system.³⁶ Hypocretin (also known as orexin) is a hormone produced in the hypothalamus that acts on multiple brain regions and maintains alertness. For unclear reasons, hypothalamic neurons producing hypocretin are selectively reduced in narcolepsy type 1. Hypocretin also stabilizes wakefulness and inhibits REM sleep; therefore, hypocretin deficiency can lead to inappropriate intrusions of REM sleep onto wakefulness, leading to the hallmark features of narcolepsy—cataplexy, sleep-related hallucinations, and sleep paralysis.³⁷ According to one theory, cataplexy is triggered by emotional stimuli because of a pathway between the medial prefrontal cortex and the amygdala to the pons.³⁸

Cerebrospinal fluid levels of hypocretin in patients with narcolepsy type 2 tend to be normal, and the biologic underpinnings of narcolepsy type 2 remain mysterious. However, in the subgroup of those with narcolepsy type 2 in which hypocretin is low, many individuals go on to develop cataplexy, thereby evolving to narcolepsy type 1.³⁶

POSSIBLE AUTOIMMUNE BASIS

Narcolepsy is typically a sporadic disorder, although familial cases have been described. The risk of a parent with narcolepsy having a child who is affected is approximately 1%.⁵

Narcolepsy type 1 is strongly associated with HLA-DQB1*0602, with up to 95% of those affected having at least one allele.³⁹ Having 2 copies of the allele further increases the risk of developing narcolepsy.⁴⁰ However, this allele is far from specific for narcolepsy with cataplexy, as it occurs in 12% to 38% of the general population.⁴¹ Therefore, HLA typing currently has limited clinical utility. The exact cause is as yet unknown, but substantial literature proposes an autoimmune basis of the disorder, given the strong association with the HLA subtype.^42–44

After the 2009 H1N1 influenza pandemic, there was a significant increase in the incidence of narcolepsy with cataplexy, which again sparked interest in an autoimmune etiology underlying the disorder. Pandemrix, an H1N1 vaccine produced as a result of the 2009 pandemic, appeared to also be associated with an increase in the incidence of narcolepsy. An association with other upper respiratory infections has also been noted, further supporting a possible autoimmune basis.

A few studies have looked for serum autoantibodies involved in the pathogenesis of narcolepsy. Thus far, only one has been identified, an antibody to Tribbles homolog 2, found in 20% to 40% of those with new onset of nar-colepsy.^42–44

TREATMENTS FOR DAYTIME SLEEPINESS

As with many chronic disorders, the treatment of narcolepsy consists of symptomatic rather than curative management, which can be done through both pharmacologic and nonpharmacologic means.

Nondrug measures

Scheduled naps lasting 15 to 20 minutes can help improve alertness.⁴⁵ A consistent sleep schedule with good sleep hygiene, ensuring sufficient nightly sleep, is also important. In one study, the combination of scheduled naps and regular nocturnal sleep times reduced the level of daytime sleepiness and unintentional daytime sleep. Daytime naps were most helpful for those with the highest degree of daytime sleepiness.⁴⁵

Strategic use of caffeine can be helpful and can reduce dependence on pharmacologic treatment.

Screening should be performed routinely for other sleep disorders, such as sleep-disordered breathing, which should be treated if identified.^5,18 When being treated for other medical conditions, individuals with narcolepsy should avoid medications that can cause sedation, such as opiates or barbiturates; alcohol should be minimized or avoided.

Networking with other individuals with narcolepsy through support groups such as Narcolepsy Network can be valuable for learning coping skills and connecting with community resources. Psychological counseling for the patient, and sometimes the family, can also be useful. School-age children may need special accommodations such as schedule adjustments to allow for scheduled naps or frequent breaks to maintain alertness.

People with narcolepsy tend to function better in careers that do not require long periods of sitting, as sleepiness tends to be worse, but instead offer flexibility and require higher levels of activity that tend to combat sleepiness. They should not work as commercial drivers.¹⁸

 

 

Medications

While behavioral interventions in narcolepsy are vital, they are rarely sufficient, and drugs that promote daytime wakefulness are used as an adjunct (Table 2).⁴⁶

Realistic expectations should be established when starting, as some degree of residual sleepiness usually remains even with optimal medical therapy. Medications should be strategically scheduled to maximize alertness during necessary times such as at work or school or during driving. Patients should specifically be counseled to avoid driving if sleepy.^18,47

Modafinil is often used as a first-line therapy, given its favorable side-effect profile and low potential for abuse. Its pharmacologic action has been debated but it probably acts as a selective dopamine reuptake inhibitor. It is typically taken twice daily (upon waking and early afternoon) and is usually well tolerated.

Potential side effects include headache, nausea, dry mouth, anorexia, diarrhea, and, rarely, Stevens-Johnson syndrome. Cardiovascular side effects are minimal, making it a favorable choice in older patients.^18,48

A trial in 283 patients showed significantly lower levels of sleepiness in patients taking modafinil 200 mg or 400 mg than in a control group. Other trials have supported these findings and showed improved driving performance on modafinil.¹⁸

Notably, modafinil can increase the metabolism of oral contraceptives, thereby reducing their efficacy. Women of childbearing age should be warned about this interaction and should be transitioned to nonhormonal forms of contraception.^2,47

Armodafinil, a purified R-isomer of modafinil, has a longer half-life and requires only once-daily dosing.⁵

If modafinil or armodafinil fails to optimally manage daytime sleepiness, a traditional stimulant such as methylphenidate or an amphetamine is often used.

Methylphenidate and amphetamines primarily inhibit the reuptake and increase the release of the monoamines, mainly dopamine, and to a lesser degree serotonin and norepinephrine.

These drugs have more significant adverse effects that can involve the cardiovascular system, causing hypertension and arrhythmias. Anorexia, weight loss, and, particularly with high doses, psychosis can occur.⁴⁹

These drugs should be avoided in patients with a history of significant cardiovascular disease. Before starting stimulant therapy, a thorough cardiovascular examination should be done, often including electrocardiography to ensure there is no baseline arrhythmia.

Patients on these medications should be followed closely to ensure that blood pressure, pulse, and weight are not negatively affected.^18,50 Addiction and tolerance can develop with these drugs, and follow-up should include assessment for dependence. Some states may require prescription drug monitoring to ensure the drugs are not being abused or diverted.

Short- and long-acting formulations of both methylphenidate and amphetamines are available, and a long-acting form is often used in conjunction with a short-acting form as needed.¹⁸

Addiction and drug-seeking behavior can develop but are unusual in those taking stimulants to treat narcolepsy.⁴⁹

Follow-up

Residual daytime sleepiness can be measured subjectively through the Epworth Sleepiness Scale on follow-up. If necessary, a maintenance-of-wakefulness test can provide an objective assessment of treatment efficacy.¹⁸

As narcolepsy is a chronic disorder, treatment should evolve with time. Most medications that treat narcolepsy are categorized by the US Food and Drug Administration as pregnancy category C, as we do not have adequate studies in human pregnancies to evaluate their effects. When a patient with narcolepsy becomes pregnant, she should be counseled about the risks and benefits of remaining on therapy. Treatment should balance the risks of sleepiness with the potential risks of remaining on medications.⁵⁰ In the elderly, as cardiovascular comorbidities tend to increase, the risks and benefits of therapy should be routinely reevaluated.

For cataplexy

Sodium oxybate,^51–53 the most potent anticataplectic drug, is the sodium salt of gamma hydroxybutyrate, a metabolite of gamma-aminobutyric acid. Sodium oxybate can be prescribed in the United States, Canada, and Europe. The American Academy of Sleep Medicine recommends sodium oxybate for cataplexy, daytime sleepiness, and disrupted sleep based on 3 level-1 studies and 2 level-4 studies.⁴⁶

Sodium oxybate increases slow-wave sleep, improves sleep continuity, and often helps to mitigate daytime sleepiness. Due to its short half-life, its administration is unusual: the first dose is taken before bedtime and the second dose 2.5 to 4 hours later. Some patients set an alarm clock to take the second dose, while others awaken spontaneously to take the second dose. Most patients find that with adherence to dosing and safety instructions, sodium oxybate can serve as a highly effective form of treatment of both excessive sleepiness and cataplexy and may reduce the need for stimulant-based therapies.

The most common adverse effects are nausea, mood swings, and enuresis. Occasionally, psychosis can result and limit use of the drug. Obstructive sleep apnea can also develop or worsen.⁵² Because of its high salt content, sodium oxybate should be used with caution in those with heart failure, hypertension, or renal impairment. Its relative, gamma hydroxybutyrate, causes rapid sedation and has been notorious for illegal use as a date rape drug.

In the United States, sodium oxybate is distributed only through a central pharmacy to mitigate potential abuse. Due to this system, the rates of diversion are extremely low, estimated in a postmarketing analysis to be 1 instance per 5,200 patients treated. In the same study, abuse and dependence were both rare as well, about 1 case for every 2,600 and 6,500 patients treated.^6,18,52,53

Antidepressants promote the action of norepinephrine and, to a lesser degree, serotonin, thereby suppressing REM sleep.

Venlafaxine, a serotonin-norepinephrine reuptake inhibitor, is often used as a first-line treatment for cataplexy. Selective serotonin reuptake inhibitors such as fluoxetine are also used with success. Tricyclic antidepressants such as protriptyline or clomipramine are extremely effective for cataplexy, but are rarely used due to their adverse effects.^2,47

FUTURE WORK

While our understanding of narcolepsy has advanced, there are still gaps in our knowledge of the disorder—namely, the specific trigger for the loss of hypocretin neurons in type 1 narcolepsy and the underlying pathophysiology of type 2.

A number of emerging therapies target the hypocretin system, including peptide replacement, neuronal transplant, and immunotherapy preventing hypocretin neuronal cell death.^50,54,55 Additional drugs designed to improve alertness that do not involve the hypocretin system are also being developed, including a histamine inverse agonist.^50,56 Sodium oxybate and modafinil, although currently approved for use in adults, are still off-label in pediatric practice. Studies of the safety and efficacy of these medications in children are needed.^7,57

Narcolepsy was originally described in the late 1800s by the French physician Jean-Baptiste-Edouard Gélineau, who reported the case of a wine merchant suffering from somnolence. In this first description, he coined the term narcolepsie by joining the Greek words narke (numbness or stupor) and lepsis (attack).¹

Since then, the disorder has been further characterized, and some insight into its biological underpinnings has been established. Importantly, treatments have improved and expanded, facilitating its management and thereby improving quality of life for those with the disorder.

This review focuses on clinically relevant features of the disorder and proposes management strategies.

CLINICAL FEATURES

Narcolepsy is characterized by instability of sleep-wake transitions.

Daytime sleepiness

Clinically, narcolepsy manifests with excessive daytime sleepiness that can be personally and socially disabling. Cataplexy, sleep paralysis, and hypnagogic or hypnopompic hallucinations can also be present,^2,3 but they are not necessary for diagnosis. In fact, a minority of patients with narcolepsy have all these symptoms.⁴ Narcolepsy is divided into type 1 (with cataplexy) and type 2 (without cataplexy).²

Sleepiness tends to be worse with inactivity, and sleep can often be irresistible. Sleep attacks can come on suddenly and may be brief enough to manifest as a lapse in consciousness.

Short naps tend to be refreshing. Rapid eye movement (REM) latency—the interval between falling asleep and the onset of the REM sleep—is short in narcolepsy, and since the REM stage is when dreaming occurs, naps often include dreaming. Therefore, when taking a history, it is worthwhile to ask patients whether they dream during naps; a yes answer supports the diagnosis of narcolepsy.⁵

In children, sleepiness can manifest in reduced concentration and behavioral issues.⁶ Napping after age 5 or 6 is considered abnormal and may reflect pathologic sleepiness.⁷

Cataplexy

Cataplexy—transient muscle weakness triggered by emotion—is a specific feature of narcolepsy type 1. It often begins in the facial muscles and can manifest with slackening of the jaw or brief dropping of the head. However, episodes can be more dramatic and, if the trunk and limb muscles are affected, can result in collapsing to the ground.

Cataplexy usually has its onset at about the same time as the sleepiness associated with narcolepsy, but it can arise even years later.⁸ Episodes can last from a few seconds to 2 minutes. Consciousness is always preserved. A range of emotions can trigger cataplexy, but typically the emotion is a positive one such as laughter or excitement.⁹ Deep tendon reflexes disappear in cataplexy, so checking reflexes during a witnessed episode can be clinically valuable.²

Cataplexy can worsen with stress and insufficient sleep, occasionally with “status cataplecticus,” in which repeated, persistent episodes of cataplexy occur over several hours.⁸ Status cataplecticus can be spontaneous or an effect of withdrawal from anticataplectic medications.²

Cataplexy is thought to represent intrusion of REM sleep and its associated muscle atonia during wakefulness.

Sleep paralysis, hallucinations

Sleep paralysis and hallucinations are other features of narcolepsy that reflect this REM dissociation from sleep.

Sleep paralysis occurs most commonly upon awakening, but sometimes just before sleep onset. In most cases, it is manifested by inability to move the limbs or speak, lasting several seconds or, in rare cases, minutes at a time. Sleep paralysis can be associated with a sensation of fear or suffocation, especially when initially experienced.⁸

Hypnopompic hallucinations, occurring upon awakening, are more common than hypnagogic hallucinations, which are experienced before falling asleep. The hallucinations are often vivid and usually visual, although other types of hallucinations are possible. Unlike those that occur in psychotic disorders, the hallucinations tend to be associated with preserved insight that they are not real.¹⁰

Notably, both sleep paralysis and hallucinations are nonspecific symptoms that are common in the general population.^8,11,12

Fragmented sleep

Although they are very sleepy, people with narcolepsy generally cannot stay asleep for very long. Their sleep tends to be extremely fragmented, and they often wake up several times a night.²

This sleep pattern reflects the inherent instability of sleep-wake transitions in narcolepsy. In fact, over a 24-hour period, adults with narcolepsy have a normal amount of sleep.¹³ In children, however, when narcolepsy first arises, the 24-hour sleep time can increase abruptly and can sometimes be associated with persistent cataplexy that can manifest as a clumsy gait.¹⁴

Weight gain, obstructive sleep apnea

Weight gain is common, particularly after symptom onset, and especially in children. As a result, obesity is a frequent comorbidity.¹⁵ Because obstructive sleep apnea can consequently develop, all patients with narcolepsy require screening for sleep-disordered breathing.

Other sleep disorders often accompany narcolepsy and are more common than in the general population.¹⁶ In a study incorporating both clinical and polysomnographic data of 100 patients with narcolepsy, insomnia was the most common comorbid sleep disorder, with a prevalence of 28%; others were REM sleep behavior disorder (24%), restless legs syndrome (24%), obstructive sleep apnea (21%), and non-REM parasomnias.¹⁷

 

 

PSYCHOSOCIAL CONSEQUENCES

Narcolepsy has significant psychosocial consequences. As a result of their symptoms, people with narcolepsy may not be able to meet academic or work-related demands.

Additionally, their risk of a motor vehicle accident is 3 to 4 times higher than in the general population, and more than one-third of patients have been in an accident due to sleepiness.¹⁸ There is some evidence to show that treatment eliminates this risk.¹⁹

Few systematic studies have examined mood disorders in narcolepsy. However, studies tend to show a higher prevalence of psychiatric disorders than in the general population, with depression and anxiety the most com-mon.^20,21

DIAGNOSIS IS OFTEN DELAYED

The prevalence of narcolepsy type 1 is between 25 and 100 per 100,000 people.²² In a Mayo Clinic study,²³ the incidence of narcolepsy type 1 was estimated to be 0.74 per 100,000 person-years. Epidemiologic data on narcolepsy type 2 are sparse, but patients with narcolepsy without cataplexy are thought to represent only 36% of all narcolepsy patients.²³

Diagnosis is often delayed, with the average time between the onset of symptoms and the diagnosis ranging from 8 to 22 years. With increasing awareness, the efficiency of the diagnostic process is improving, and this delay is expected to lessen accordingly.²⁴

Symptoms most commonly arise in the second decade; but the age at onset ranges significantly, between the first and fifth decades. Narcolepsy has a bimodal distribution in incidence, with the biggest peak at approximately age 15 and second smaller peak in the mid-30s. Some studies have suggested a slight male predominance.^23,25

DIAGNOSIS

History is key

The history should include specific questions about the hallmark features of narcolepsy, including cataplexy, sleep paralysis, and sleep-related hallucinations. For individual assessment of subjective sleepiness, the Epworth Sleepiness Scale or Pediatric Daytime Sleepiness Scale can be administered quickly in the office setting.^26,27

The Epworth score is calculated from the self-rated likelihood of falling asleep in 8 different situations, with possible scores of 0 (would never doze) to 3 (high chance of dozing) on each question, for a total possible score of 0 to 24. Normal total scores are between 0 and 10, while scores greater than 10 reflect pathologic sleepiness. Scores on the Epworth Sleepiness Scale in those with narcolepsy tend to reflect moderate to severe sleepiness, or at least 13, as opposed to patients with obstructive sleep apnea, whose scores commonly reflect milder sleepiness.²⁸

Testing with actigraphy and polysomnography

It is imperative to rule out insufficient sleep and other sleep disorders as a cause of daytime sleepiness. This can be done with a careful clinical history, actigraphy with sleep logs, and polysomnography.

In the 2 to 4 weeks before actigraphy and subsequent testing, all medications with alerting or sedating properties (including anti­depressants) should be tapered off to prevent influence on the results of the study.

Delayed sleep-phase disorder presents at a similar age as narcolepsy and can be associated with similar degrees of sleepiness. However, individuals with delayed sleep phase disorder have an inappropriately timed sleep-wake cycle so that there is a shift in their desired sleep onset and awakening times. It is common—prevalence estimates vary but average about 1% in the general population.²⁹

Insufficient sleep syndrome is even more common, especially in teenagers and young adults, with increasing family, social, and academic demands. Sleep needs vary across the life span. A teenager needs 8 to 10 hours of sleep per night, and a young adult needs 7 to 9 hours. A study of 1,285 high school students found that 10.4% were not getting enough sleep.³⁰

If actigraphy data suggest a circadian rhythm disorder or insufficient sleep that could explain the symptoms of sleepiness, then further testing should be halted and these specific issues should be addressed. In these cases, working with the patient toward maintaining a regular sleep-wake schedule with 7 to 8 hours of nightly sleep will often resolve symptoms.

If actigraphy demonstrates the patient is maintaining a regular sleep schedule and allowing adequate time for nightly sleep, the next step is polysomnography.

Polysomnography is performed to detect other disorders that can disrupt sleep, such as sleep-disordered breathing or periodic limb movement disorder.^2,5 In addition, polysomnography can provide assurance that adequate sleep was obtained prior to the next step in testing.

Multiple sleep latency test

If sufficient sleep is obtained on polysomnograpy (at least 6 hours for an adult) and no other sleep disorder is identified, a multiple sleep latency test is performed. A urine toxicology screen is typically performed on the day of the test to ensure that drugs are not affecting the results.

The multiple sleep latency test consists of 4 to 5 nap opportunities at 2-hour intervals in a quiet dark room conducive to sleep, during which both sleep and REM latency are recorded. The sleep latency of those with narcolepsy is significantly shortened, and the diagnosis of narcolepsy requires an average sleep latency of less than 8 minutes.

Given the propensity for REM sleep in narcolepsy, another essential feature for diagnosis is the sleep-onset REM period (SOREMP). A SOREMP is defined as a REM latency of less than 15 minutes. A diagnosis of narcolepsy re-quires a SOREMP in at least 2 of the naps in a multiple sleep latency test (or 1 nap if the shortened REM latency is seen during polysomnography).³¹

The multiple sleep latency test has an imperfect sensitivity, though, and should be repeated when there is a high suspicion of narcolepsy.^32–34 It is not completely specific either, and false-positive results occur. In fact, SOREMPs can be seen in the general population, particularly in those with a circadian rhythm disorder, insufficient sleep, or sleep-disordered breathing. Two or more SOREMPs in an multiple sleep latency test can be seen in a small proportion of the general population.³⁵ The results of a multiple sleep latency test should be interpreted in the clinical context.

Differential diagnosis

Narcolepsy type 1 is distinguished from type 2 by the presence of cataplexy. A cerebrospinal fluid hypocretin 1 level of 110 pg/mL or less, or less than one-third of the mean value obtained in normal individuals, can substitute for the multiple sleep latency test in diagnosing narcolepsy type 1.³¹ Currently, hypocretin testing is generally not performed in clinical practice, although it may become a routine part of the narcolepsy evaluation in the future.

Thus, according to the International Classification of Sleep Disorders, 3rd edition,³¹ the diagnosis of narcolepsy type 1 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurologic disorder, mental disorder, medication use, or substance use disorder, and at least 1 of the following:

• Cataplexy and mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing (1 of which can be on the preceding night’s polysomography)
• Cerebrospinal fluid hypocretin 1 levels less than 110 pg/mL or one-third the baseline normal levels and mean sleep latency ≤ 8 minutes with ≥ 2 SOREMPs on multiple sleep latency testing.

Similarly, the diagnosis of narcolepsy type 2 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurological disorder, mental disorder, medication use, or substance use disorder, plus:

• Mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing.

Idiopathic hypersomnia, another disorder of central hypersomnolence, is also characterized by disabling sleepiness. It is diagnostically differentiated from narcolepsy, as there are fewer than 2 SOREMPs. As opposed to narcolepsy, in which naps tend to be refreshing, even prolonged naps in idiopathic hypersomnia are often not helpful in restoring wakefulness. In idiopathic hypersomnia, sleep is usually not fragmented, and there are few nocturnal arousals. Sleep times can often be prolonged as well, whereas in narcolepsy total sleep time through the day may not be increased but is not consolidated.

Kleine-Levin syndrome is a rarer disorder of hypersomnia. It is episodic compared with the relatively persistent sleepiness in narcolepsy and idiopathic hypersomnia. Periods of hypersomnia occur intermittently for days to weeks and are accompanied by cognitive and behavioral changes including hyperphagia and hypersexuality.⁴

LINKED TO HYPOCRETIN DEFICIENCY

Over the past 2 decades, the underlying pathophysiology of narcolepsy type 1 has been better characterized.

Narcolepsy type 1 has been linked to a deficiency in hypocretin in the central nervous system.³⁶ Hypocretin (also known as orexin) is a hormone produced in the hypothalamus that acts on multiple brain regions and maintains alertness. For unclear reasons, hypothalamic neurons producing hypocretin are selectively reduced in narcolepsy type 1. Hypocretin also stabilizes wakefulness and inhibits REM sleep; therefore, hypocretin deficiency can lead to inappropriate intrusions of REM sleep onto wakefulness, leading to the hallmark features of narcolepsy—cataplexy, sleep-related hallucinations, and sleep paralysis.³⁷ According to one theory, cataplexy is triggered by emotional stimuli because of a pathway between the medial prefrontal cortex and the amygdala to the pons.³⁸

Cerebrospinal fluid levels of hypocretin in patients with narcolepsy type 2 tend to be normal, and the biologic underpinnings of narcolepsy type 2 remain mysterious. However, in the subgroup of those with narcolepsy type 2 in which hypocretin is low, many individuals go on to develop cataplexy, thereby evolving to narcolepsy type 1.³⁶

POSSIBLE AUTOIMMUNE BASIS

Narcolepsy is typically a sporadic disorder, although familial cases have been described. The risk of a parent with narcolepsy having a child who is affected is approximately 1%.⁵

Narcolepsy type 1 is strongly associated with HLA-DQB1*0602, with up to 95% of those affected having at least one allele.³⁹ Having 2 copies of the allele further increases the risk of developing narcolepsy.⁴⁰ However, this allele is far from specific for narcolepsy with cataplexy, as it occurs in 12% to 38% of the general population.⁴¹ Therefore, HLA typing currently has limited clinical utility. The exact cause is as yet unknown, but substantial literature proposes an autoimmune basis of the disorder, given the strong association with the HLA subtype.^42–44

After the 2009 H1N1 influenza pandemic, there was a significant increase in the incidence of narcolepsy with cataplexy, which again sparked interest in an autoimmune etiology underlying the disorder. Pandemrix, an H1N1 vaccine produced as a result of the 2009 pandemic, appeared to also be associated with an increase in the incidence of narcolepsy. An association with other upper respiratory infections has also been noted, further supporting a possible autoimmune basis.

A few studies have looked for serum autoantibodies involved in the pathogenesis of narcolepsy. Thus far, only one has been identified, an antibody to Tribbles homolog 2, found in 20% to 40% of those with new onset of nar-colepsy.^42–44

TREATMENTS FOR DAYTIME SLEEPINESS

As with many chronic disorders, the treatment of narcolepsy consists of symptomatic rather than curative management, which can be done through both pharmacologic and nonpharmacologic means.

Nondrug measures

Scheduled naps lasting 15 to 20 minutes can help improve alertness.⁴⁵ A consistent sleep schedule with good sleep hygiene, ensuring sufficient nightly sleep, is also important. In one study, the combination of scheduled naps and regular nocturnal sleep times reduced the level of daytime sleepiness and unintentional daytime sleep. Daytime naps were most helpful for those with the highest degree of daytime sleepiness.⁴⁵

Strategic use of caffeine can be helpful and can reduce dependence on pharmacologic treatment.

Screening should be performed routinely for other sleep disorders, such as sleep-disordered breathing, which should be treated if identified.^5,18 When being treated for other medical conditions, individuals with narcolepsy should avoid medications that can cause sedation, such as opiates or barbiturates; alcohol should be minimized or avoided.

Networking with other individuals with narcolepsy through support groups such as Narcolepsy Network can be valuable for learning coping skills and connecting with community resources. Psychological counseling for the patient, and sometimes the family, can also be useful. School-age children may need special accommodations such as schedule adjustments to allow for scheduled naps or frequent breaks to maintain alertness.

People with narcolepsy tend to function better in careers that do not require long periods of sitting, as sleepiness tends to be worse, but instead offer flexibility and require higher levels of activity that tend to combat sleepiness. They should not work as commercial drivers.¹⁸

 

 

Medications

While behavioral interventions in narcolepsy are vital, they are rarely sufficient, and drugs that promote daytime wakefulness are used as an adjunct (Table 2).⁴⁶

Realistic expectations should be established when starting, as some degree of residual sleepiness usually remains even with optimal medical therapy. Medications should be strategically scheduled to maximize alertness during necessary times such as at work or school or during driving. Patients should specifically be counseled to avoid driving if sleepy.^18,47

Modafinil is often used as a first-line therapy, given its favorable side-effect profile and low potential for abuse. Its pharmacologic action has been debated but it probably acts as a selective dopamine reuptake inhibitor. It is typically taken twice daily (upon waking and early afternoon) and is usually well tolerated.

Potential side effects include headache, nausea, dry mouth, anorexia, diarrhea, and, rarely, Stevens-Johnson syndrome. Cardiovascular side effects are minimal, making it a favorable choice in older patients.^18,48

A trial in 283 patients showed significantly lower levels of sleepiness in patients taking modafinil 200 mg or 400 mg than in a control group. Other trials have supported these findings and showed improved driving performance on modafinil.¹⁸

Notably, modafinil can increase the metabolism of oral contraceptives, thereby reducing their efficacy. Women of childbearing age should be warned about this interaction and should be transitioned to nonhormonal forms of contraception.^2,47

Armodafinil, a purified R-isomer of modafinil, has a longer half-life and requires only once-daily dosing.⁵

If modafinil or armodafinil fails to optimally manage daytime sleepiness, a traditional stimulant such as methylphenidate or an amphetamine is often used.

Methylphenidate and amphetamines primarily inhibit the reuptake and increase the release of the monoamines, mainly dopamine, and to a lesser degree serotonin and norepinephrine.

These drugs have more significant adverse effects that can involve the cardiovascular system, causing hypertension and arrhythmias. Anorexia, weight loss, and, particularly with high doses, psychosis can occur.⁴⁹

These drugs should be avoided in patients with a history of significant cardiovascular disease. Before starting stimulant therapy, a thorough cardiovascular examination should be done, often including electrocardiography to ensure there is no baseline arrhythmia.

Patients on these medications should be followed closely to ensure that blood pressure, pulse, and weight are not negatively affected.^18,50 Addiction and tolerance can develop with these drugs, and follow-up should include assessment for dependence. Some states may require prescription drug monitoring to ensure the drugs are not being abused or diverted.

Short- and long-acting formulations of both methylphenidate and amphetamines are available, and a long-acting form is often used in conjunction with a short-acting form as needed.¹⁸

Addiction and drug-seeking behavior can develop but are unusual in those taking stimulants to treat narcolepsy.⁴⁹

Follow-up

Residual daytime sleepiness can be measured subjectively through the Epworth Sleepiness Scale on follow-up. If necessary, a maintenance-of-wakefulness test can provide an objective assessment of treatment efficacy.¹⁸

As narcolepsy is a chronic disorder, treatment should evolve with time. Most medications that treat narcolepsy are categorized by the US Food and Drug Administration as pregnancy category C, as we do not have adequate studies in human pregnancies to evaluate their effects. When a patient with narcolepsy becomes pregnant, she should be counseled about the risks and benefits of remaining on therapy. Treatment should balance the risks of sleepiness with the potential risks of remaining on medications.⁵⁰ In the elderly, as cardiovascular comorbidities tend to increase, the risks and benefits of therapy should be routinely reevaluated.

For cataplexy

Sodium oxybate,^51–53 the most potent anticataplectic drug, is the sodium salt of gamma hydroxybutyrate, a metabolite of gamma-aminobutyric acid. Sodium oxybate can be prescribed in the United States, Canada, and Europe. The American Academy of Sleep Medicine recommends sodium oxybate for cataplexy, daytime sleepiness, and disrupted sleep based on 3 level-1 studies and 2 level-4 studies.⁴⁶

Sodium oxybate increases slow-wave sleep, improves sleep continuity, and often helps to mitigate daytime sleepiness. Due to its short half-life, its administration is unusual: the first dose is taken before bedtime and the second dose 2.5 to 4 hours later. Some patients set an alarm clock to take the second dose, while others awaken spontaneously to take the second dose. Most patients find that with adherence to dosing and safety instructions, sodium oxybate can serve as a highly effective form of treatment of both excessive sleepiness and cataplexy and may reduce the need for stimulant-based therapies.

The most common adverse effects are nausea, mood swings, and enuresis. Occasionally, psychosis can result and limit use of the drug. Obstructive sleep apnea can also develop or worsen.⁵² Because of its high salt content, sodium oxybate should be used with caution in those with heart failure, hypertension, or renal impairment. Its relative, gamma hydroxybutyrate, causes rapid sedation and has been notorious for illegal use as a date rape drug.

In the United States, sodium oxybate is distributed only through a central pharmacy to mitigate potential abuse. Due to this system, the rates of diversion are extremely low, estimated in a postmarketing analysis to be 1 instance per 5,200 patients treated. In the same study, abuse and dependence were both rare as well, about 1 case for every 2,600 and 6,500 patients treated.^6,18,52,53

Antidepressants promote the action of norepinephrine and, to a lesser degree, serotonin, thereby suppressing REM sleep.

Venlafaxine, a serotonin-norepinephrine reuptake inhibitor, is often used as a first-line treatment for cataplexy. Selective serotonin reuptake inhibitors such as fluoxetine are also used with success. Tricyclic antidepressants such as protriptyline or clomipramine are extremely effective for cataplexy, but are rarely used due to their adverse effects.^2,47

FUTURE WORK

While our understanding of narcolepsy has advanced, there are still gaps in our knowledge of the disorder—namely, the specific trigger for the loss of hypocretin neurons in type 1 narcolepsy and the underlying pathophysiology of type 2.

A number of emerging therapies target the hypocretin system, including peptide replacement, neuronal transplant, and immunotherapy preventing hypocretin neuronal cell death.^50,54,55 Additional drugs designed to improve alertness that do not involve the hypocretin system are also being developed, including a histamine inverse agonist.^50,56 Sodium oxybate and modafinil, although currently approved for use in adults, are still off-label in pediatric practice. Studies of the safety and efficacy of these medications in children are needed.^7,57

Narcolepsy was originally described in the late 1800s by the French physician Jean-Baptiste-Edouard Gélineau, who reported the case of a wine merchant suffering from somnolence. In this first description, he coined the term narcolepsie by joining the Greek words narke (numbness or stupor) and lepsis (attack).¹

Since then, the disorder has been further characterized, and some insight into its biological underpinnings has been established. Importantly, treatments have improved and expanded, facilitating its management and thereby improving quality of life for those with the disorder.

This review focuses on clinically relevant features of the disorder and proposes management strategies.

CLINICAL FEATURES

Narcolepsy is characterized by instability of sleep-wake transitions.

Daytime sleepiness

Clinically, narcolepsy manifests with excessive daytime sleepiness that can be personally and socially disabling. Cataplexy, sleep paralysis, and hypnagogic or hypnopompic hallucinations can also be present,^2,3 but they are not necessary for diagnosis. In fact, a minority of patients with narcolepsy have all these symptoms.⁴ Narcolepsy is divided into type 1 (with cataplexy) and type 2 (without cataplexy).²

Sleepiness tends to be worse with inactivity, and sleep can often be irresistible. Sleep attacks can come on suddenly and may be brief enough to manifest as a lapse in consciousness.

Short naps tend to be refreshing. Rapid eye movement (REM) latency—the interval between falling asleep and the onset of the REM sleep—is short in narcolepsy, and since the REM stage is when dreaming occurs, naps often include dreaming. Therefore, when taking a history, it is worthwhile to ask patients whether they dream during naps; a yes answer supports the diagnosis of narcolepsy.⁵

In children, sleepiness can manifest in reduced concentration and behavioral issues.⁶ Napping after age 5 or 6 is considered abnormal and may reflect pathologic sleepiness.⁷

Cataplexy

Cataplexy—transient muscle weakness triggered by emotion—is a specific feature of narcolepsy type 1. It often begins in the facial muscles and can manifest with slackening of the jaw or brief dropping of the head. However, episodes can be more dramatic and, if the trunk and limb muscles are affected, can result in collapsing to the ground.

Cataplexy usually has its onset at about the same time as the sleepiness associated with narcolepsy, but it can arise even years later.⁸ Episodes can last from a few seconds to 2 minutes. Consciousness is always preserved. A range of emotions can trigger cataplexy, but typically the emotion is a positive one such as laughter or excitement.⁹ Deep tendon reflexes disappear in cataplexy, so checking reflexes during a witnessed episode can be clinically valuable.²

Cataplexy can worsen with stress and insufficient sleep, occasionally with “status cataplecticus,” in which repeated, persistent episodes of cataplexy occur over several hours.⁸ Status cataplecticus can be spontaneous or an effect of withdrawal from anticataplectic medications.²

Cataplexy is thought to represent intrusion of REM sleep and its associated muscle atonia during wakefulness.

Sleep paralysis, hallucinations

Sleep paralysis and hallucinations are other features of narcolepsy that reflect this REM dissociation from sleep.

Sleep paralysis occurs most commonly upon awakening, but sometimes just before sleep onset. In most cases, it is manifested by inability to move the limbs or speak, lasting several seconds or, in rare cases, minutes at a time. Sleep paralysis can be associated with a sensation of fear or suffocation, especially when initially experienced.⁸

Hypnopompic hallucinations, occurring upon awakening, are more common than hypnagogic hallucinations, which are experienced before falling asleep. The hallucinations are often vivid and usually visual, although other types of hallucinations are possible. Unlike those that occur in psychotic disorders, the hallucinations tend to be associated with preserved insight that they are not real.¹⁰

Notably, both sleep paralysis and hallucinations are nonspecific symptoms that are common in the general population.^8,11,12

Fragmented sleep

Although they are very sleepy, people with narcolepsy generally cannot stay asleep for very long. Their sleep tends to be extremely fragmented, and they often wake up several times a night.²

This sleep pattern reflects the inherent instability of sleep-wake transitions in narcolepsy. In fact, over a 24-hour period, adults with narcolepsy have a normal amount of sleep.¹³ In children, however, when narcolepsy first arises, the 24-hour sleep time can increase abruptly and can sometimes be associated with persistent cataplexy that can manifest as a clumsy gait.¹⁴

Weight gain, obstructive sleep apnea

Weight gain is common, particularly after symptom onset, and especially in children. As a result, obesity is a frequent comorbidity.¹⁵ Because obstructive sleep apnea can consequently develop, all patients with narcolepsy require screening for sleep-disordered breathing.

Other sleep disorders often accompany narcolepsy and are more common than in the general population.¹⁶ In a study incorporating both clinical and polysomnographic data of 100 patients with narcolepsy, insomnia was the most common comorbid sleep disorder, with a prevalence of 28%; others were REM sleep behavior disorder (24%), restless legs syndrome (24%), obstructive sleep apnea (21%), and non-REM parasomnias.¹⁷

 

 

PSYCHOSOCIAL CONSEQUENCES

Narcolepsy has significant psychosocial consequences. As a result of their symptoms, people with narcolepsy may not be able to meet academic or work-related demands.

Additionally, their risk of a motor vehicle accident is 3 to 4 times higher than in the general population, and more than one-third of patients have been in an accident due to sleepiness.¹⁸ There is some evidence to show that treatment eliminates this risk.¹⁹

Few systematic studies have examined mood disorders in narcolepsy. However, studies tend to show a higher prevalence of psychiatric disorders than in the general population, with depression and anxiety the most com-mon.^20,21

DIAGNOSIS IS OFTEN DELAYED

The prevalence of narcolepsy type 1 is between 25 and 100 per 100,000 people.²² In a Mayo Clinic study,²³ the incidence of narcolepsy type 1 was estimated to be 0.74 per 100,000 person-years. Epidemiologic data on narcolepsy type 2 are sparse, but patients with narcolepsy without cataplexy are thought to represent only 36% of all narcolepsy patients.²³

Diagnosis is often delayed, with the average time between the onset of symptoms and the diagnosis ranging from 8 to 22 years. With increasing awareness, the efficiency of the diagnostic process is improving, and this delay is expected to lessen accordingly.²⁴

Symptoms most commonly arise in the second decade; but the age at onset ranges significantly, between the first and fifth decades. Narcolepsy has a bimodal distribution in incidence, with the biggest peak at approximately age 15 and second smaller peak in the mid-30s. Some studies have suggested a slight male predominance.^23,25

DIAGNOSIS

History is key

The history should include specific questions about the hallmark features of narcolepsy, including cataplexy, sleep paralysis, and sleep-related hallucinations. For individual assessment of subjective sleepiness, the Epworth Sleepiness Scale or Pediatric Daytime Sleepiness Scale can be administered quickly in the office setting.^26,27

The Epworth score is calculated from the self-rated likelihood of falling asleep in 8 different situations, with possible scores of 0 (would never doze) to 3 (high chance of dozing) on each question, for a total possible score of 0 to 24. Normal total scores are between 0 and 10, while scores greater than 10 reflect pathologic sleepiness. Scores on the Epworth Sleepiness Scale in those with narcolepsy tend to reflect moderate to severe sleepiness, or at least 13, as opposed to patients with obstructive sleep apnea, whose scores commonly reflect milder sleepiness.²⁸

Testing with actigraphy and polysomnography

It is imperative to rule out insufficient sleep and other sleep disorders as a cause of daytime sleepiness. This can be done with a careful clinical history, actigraphy with sleep logs, and polysomnography.

In the 2 to 4 weeks before actigraphy and subsequent testing, all medications with alerting or sedating properties (including anti­depressants) should be tapered off to prevent influence on the results of the study.

Delayed sleep-phase disorder presents at a similar age as narcolepsy and can be associated with similar degrees of sleepiness. However, individuals with delayed sleep phase disorder have an inappropriately timed sleep-wake cycle so that there is a shift in their desired sleep onset and awakening times. It is common—prevalence estimates vary but average about 1% in the general population.²⁹

Insufficient sleep syndrome is even more common, especially in teenagers and young adults, with increasing family, social, and academic demands. Sleep needs vary across the life span. A teenager needs 8 to 10 hours of sleep per night, and a young adult needs 7 to 9 hours. A study of 1,285 high school students found that 10.4% were not getting enough sleep.³⁰

If actigraphy data suggest a circadian rhythm disorder or insufficient sleep that could explain the symptoms of sleepiness, then further testing should be halted and these specific issues should be addressed. In these cases, working with the patient toward maintaining a regular sleep-wake schedule with 7 to 8 hours of nightly sleep will often resolve symptoms.

If actigraphy demonstrates the patient is maintaining a regular sleep schedule and allowing adequate time for nightly sleep, the next step is polysomnography.

Polysomnography is performed to detect other disorders that can disrupt sleep, such as sleep-disordered breathing or periodic limb movement disorder.^2,5 In addition, polysomnography can provide assurance that adequate sleep was obtained prior to the next step in testing.

Multiple sleep latency test

If sufficient sleep is obtained on polysomnograpy (at least 6 hours for an adult) and no other sleep disorder is identified, a multiple sleep latency test is performed. A urine toxicology screen is typically performed on the day of the test to ensure that drugs are not affecting the results.

The multiple sleep latency test consists of 4 to 5 nap opportunities at 2-hour intervals in a quiet dark room conducive to sleep, during which both sleep and REM latency are recorded. The sleep latency of those with narcolepsy is significantly shortened, and the diagnosis of narcolepsy requires an average sleep latency of less than 8 minutes.

Given the propensity for REM sleep in narcolepsy, another essential feature for diagnosis is the sleep-onset REM period (SOREMP). A SOREMP is defined as a REM latency of less than 15 minutes. A diagnosis of narcolepsy re-quires a SOREMP in at least 2 of the naps in a multiple sleep latency test (or 1 nap if the shortened REM latency is seen during polysomnography).³¹

The multiple sleep latency test has an imperfect sensitivity, though, and should be repeated when there is a high suspicion of narcolepsy.^32–34 It is not completely specific either, and false-positive results occur. In fact, SOREMPs can be seen in the general population, particularly in those with a circadian rhythm disorder, insufficient sleep, or sleep-disordered breathing. Two or more SOREMPs in an multiple sleep latency test can be seen in a small proportion of the general population.³⁵ The results of a multiple sleep latency test should be interpreted in the clinical context.

Differential diagnosis

Narcolepsy type 1 is distinguished from type 2 by the presence of cataplexy. A cerebrospinal fluid hypocretin 1 level of 110 pg/mL or less, or less than one-third of the mean value obtained in normal individuals, can substitute for the multiple sleep latency test in diagnosing narcolepsy type 1.³¹ Currently, hypocretin testing is generally not performed in clinical practice, although it may become a routine part of the narcolepsy evaluation in the future.

Thus, according to the International Classification of Sleep Disorders, 3rd edition,³¹ the diagnosis of narcolepsy type 1 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurologic disorder, mental disorder, medication use, or substance use disorder, and at least 1 of the following:

• Cataplexy and mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing (1 of which can be on the preceding night’s polysomography)
• Cerebrospinal fluid hypocretin 1 levels less than 110 pg/mL or one-third the baseline normal levels and mean sleep latency ≤ 8 minutes with ≥ 2 SOREMPs on multiple sleep latency testing.

Similarly, the diagnosis of narcolepsy type 2 requires excessive daytime sleepiness for at least 3 months that cannot be explained by another sleep disorder, medical or neurological disorder, mental disorder, medication use, or substance use disorder, plus:

• Mean sleep latency of 8 minutes or less with at least 2 SOREMPs on multiple sleep latency testing.

Idiopathic hypersomnia, another disorder of central hypersomnolence, is also characterized by disabling sleepiness. It is diagnostically differentiated from narcolepsy, as there are fewer than 2 SOREMPs. As opposed to narcolepsy, in which naps tend to be refreshing, even prolonged naps in idiopathic hypersomnia are often not helpful in restoring wakefulness. In idiopathic hypersomnia, sleep is usually not fragmented, and there are few nocturnal arousals. Sleep times can often be prolonged as well, whereas in narcolepsy total sleep time through the day may not be increased but is not consolidated.

Kleine-Levin syndrome is a rarer disorder of hypersomnia. It is episodic compared with the relatively persistent sleepiness in narcolepsy and idiopathic hypersomnia. Periods of hypersomnia occur intermittently for days to weeks and are accompanied by cognitive and behavioral changes including hyperphagia and hypersexuality.⁴

LINKED TO HYPOCRETIN DEFICIENCY

Over the past 2 decades, the underlying pathophysiology of narcolepsy type 1 has been better characterized.

Narcolepsy type 1 has been linked to a deficiency in hypocretin in the central nervous system.³⁶ Hypocretin (also known as orexin) is a hormone produced in the hypothalamus that acts on multiple brain regions and maintains alertness. For unclear reasons, hypothalamic neurons producing hypocretin are selectively reduced in narcolepsy type 1. Hypocretin also stabilizes wakefulness and inhibits REM sleep; therefore, hypocretin deficiency can lead to inappropriate intrusions of REM sleep onto wakefulness, leading to the hallmark features of narcolepsy—cataplexy, sleep-related hallucinations, and sleep paralysis.³⁷ According to one theory, cataplexy is triggered by emotional stimuli because of a pathway between the medial prefrontal cortex and the amygdala to the pons.³⁸

Cerebrospinal fluid levels of hypocretin in patients with narcolepsy type 2 tend to be normal, and the biologic underpinnings of narcolepsy type 2 remain mysterious. However, in the subgroup of those with narcolepsy type 2 in which hypocretin is low, many individuals go on to develop cataplexy, thereby evolving to narcolepsy type 1.³⁶

POSSIBLE AUTOIMMUNE BASIS

Narcolepsy is typically a sporadic disorder, although familial cases have been described. The risk of a parent with narcolepsy having a child who is affected is approximately 1%.⁵

Narcolepsy type 1 is strongly associated with HLA-DQB1*0602, with up to 95% of those affected having at least one allele.³⁹ Having 2 copies of the allele further increases the risk of developing narcolepsy.⁴⁰ However, this allele is far from specific for narcolepsy with cataplexy, as it occurs in 12% to 38% of the general population.⁴¹ Therefore, HLA typing currently has limited clinical utility. The exact cause is as yet unknown, but substantial literature proposes an autoimmune basis of the disorder, given the strong association with the HLA subtype.^42–44

After the 2009 H1N1 influenza pandemic, there was a significant increase in the incidence of narcolepsy with cataplexy, which again sparked interest in an autoimmune etiology underlying the disorder. Pandemrix, an H1N1 vaccine produced as a result of the 2009 pandemic, appeared to also be associated with an increase in the incidence of narcolepsy. An association with other upper respiratory infections has also been noted, further supporting a possible autoimmune basis.

A few studies have looked for serum autoantibodies involved in the pathogenesis of narcolepsy. Thus far, only one has been identified, an antibody to Tribbles homolog 2, found in 20% to 40% of those with new onset of nar-colepsy.^42–44

TREATMENTS FOR DAYTIME SLEEPINESS

As with many chronic disorders, the treatment of narcolepsy consists of symptomatic rather than curative management, which can be done through both pharmacologic and nonpharmacologic means.

Nondrug measures

Scheduled naps lasting 15 to 20 minutes can help improve alertness.⁴⁵ A consistent sleep schedule with good sleep hygiene, ensuring sufficient nightly sleep, is also important. In one study, the combination of scheduled naps and regular nocturnal sleep times reduced the level of daytime sleepiness and unintentional daytime sleep. Daytime naps were most helpful for those with the highest degree of daytime sleepiness.⁴⁵

Strategic use of caffeine can be helpful and can reduce dependence on pharmacologic treatment.

Screening should be performed routinely for other sleep disorders, such as sleep-disordered breathing, which should be treated if identified.^5,18 When being treated for other medical conditions, individuals with narcolepsy should avoid medications that can cause sedation, such as opiates or barbiturates; alcohol should be minimized or avoided.

Networking with other individuals with narcolepsy through support groups such as Narcolepsy Network can be valuable for learning coping skills and connecting with community resources. Psychological counseling for the patient, and sometimes the family, can also be useful. School-age children may need special accommodations such as schedule adjustments to allow for scheduled naps or frequent breaks to maintain alertness.

People with narcolepsy tend to function better in careers that do not require long periods of sitting, as sleepiness tends to be worse, but instead offer flexibility and require higher levels of activity that tend to combat sleepiness. They should not work as commercial drivers.¹⁸

 

 

Medications

While behavioral interventions in narcolepsy are vital, they are rarely sufficient, and drugs that promote daytime wakefulness are used as an adjunct (Table 2).⁴⁶

Realistic expectations should be established when starting, as some degree of residual sleepiness usually remains even with optimal medical therapy. Medications should be strategically scheduled to maximize alertness during necessary times such as at work or school or during driving. Patients should specifically be counseled to avoid driving if sleepy.^18,47

Modafinil is often used as a first-line therapy, given its favorable side-effect profile and low potential for abuse. Its pharmacologic action has been debated but it probably acts as a selective dopamine reuptake inhibitor. It is typically taken twice daily (upon waking and early afternoon) and is usually well tolerated.

Potential side effects include headache, nausea, dry mouth, anorexia, diarrhea, and, rarely, Stevens-Johnson syndrome. Cardiovascular side effects are minimal, making it a favorable choice in older patients.^18,48

A trial in 283 patients showed significantly lower levels of sleepiness in patients taking modafinil 200 mg or 400 mg than in a control group. Other trials have supported these findings and showed improved driving performance on modafinil.¹⁸

Notably, modafinil can increase the metabolism of oral contraceptives, thereby reducing their efficacy. Women of childbearing age should be warned about this interaction and should be transitioned to nonhormonal forms of contraception.^2,47

Armodafinil, a purified R-isomer of modafinil, has a longer half-life and requires only once-daily dosing.⁵

If modafinil or armodafinil fails to optimally manage daytime sleepiness, a traditional stimulant such as methylphenidate or an amphetamine is often used.

Methylphenidate and amphetamines primarily inhibit the reuptake and increase the release of the monoamines, mainly dopamine, and to a lesser degree serotonin and norepinephrine.

These drugs have more significant adverse effects that can involve the cardiovascular system, causing hypertension and arrhythmias. Anorexia, weight loss, and, particularly with high doses, psychosis can occur.⁴⁹

These drugs should be avoided in patients with a history of significant cardiovascular disease. Before starting stimulant therapy, a thorough cardiovascular examination should be done, often including electrocardiography to ensure there is no baseline arrhythmia.

Patients on these medications should be followed closely to ensure that blood pressure, pulse, and weight are not negatively affected.^18,50 Addiction and tolerance can develop with these drugs, and follow-up should include assessment for dependence. Some states may require prescription drug monitoring to ensure the drugs are not being abused or diverted.

Short- and long-acting formulations of both methylphenidate and amphetamines are available, and a long-acting form is often used in conjunction with a short-acting form as needed.¹⁸

Addiction and drug-seeking behavior can develop but are unusual in those taking stimulants to treat narcolepsy.⁴⁹

Follow-up

Residual daytime sleepiness can be measured subjectively through the Epworth Sleepiness Scale on follow-up. If necessary, a maintenance-of-wakefulness test can provide an objective assessment of treatment efficacy.¹⁸

As narcolepsy is a chronic disorder, treatment should evolve with time. Most medications that treat narcolepsy are categorized by the US Food and Drug Administration as pregnancy category C, as we do not have adequate studies in human pregnancies to evaluate their effects. When a patient with narcolepsy becomes pregnant, she should be counseled about the risks and benefits of remaining on therapy. Treatment should balance the risks of sleepiness with the potential risks of remaining on medications.⁵⁰ In the elderly, as cardiovascular comorbidities tend to increase, the risks and benefits of therapy should be routinely reevaluated.

For cataplexy

Sodium oxybate,^51–53 the most potent anticataplectic drug, is the sodium salt of gamma hydroxybutyrate, a metabolite of gamma-aminobutyric acid. Sodium oxybate can be prescribed in the United States, Canada, and Europe. The American Academy of Sleep Medicine recommends sodium oxybate for cataplexy, daytime sleepiness, and disrupted sleep based on 3 level-1 studies and 2 level-4 studies.⁴⁶

Sodium oxybate increases slow-wave sleep, improves sleep continuity, and often helps to mitigate daytime sleepiness. Due to its short half-life, its administration is unusual: the first dose is taken before bedtime and the second dose 2.5 to 4 hours later. Some patients set an alarm clock to take the second dose, while others awaken spontaneously to take the second dose. Most patients find that with adherence to dosing and safety instructions, sodium oxybate can serve as a highly effective form of treatment of both excessive sleepiness and cataplexy and may reduce the need for stimulant-based therapies.

The most common adverse effects are nausea, mood swings, and enuresis. Occasionally, psychosis can result and limit use of the drug. Obstructive sleep apnea can also develop or worsen.⁵² Because of its high salt content, sodium oxybate should be used with caution in those with heart failure, hypertension, or renal impairment. Its relative, gamma hydroxybutyrate, causes rapid sedation and has been notorious for illegal use as a date rape drug.

In the United States, sodium oxybate is distributed only through a central pharmacy to mitigate potential abuse. Due to this system, the rates of diversion are extremely low, estimated in a postmarketing analysis to be 1 instance per 5,200 patients treated. In the same study, abuse and dependence were both rare as well, about 1 case for every 2,600 and 6,500 patients treated.^6,18,52,53

Antidepressants promote the action of norepinephrine and, to a lesser degree, serotonin, thereby suppressing REM sleep.

Venlafaxine, a serotonin-norepinephrine reuptake inhibitor, is often used as a first-line treatment for cataplexy. Selective serotonin reuptake inhibitors such as fluoxetine are also used with success. Tricyclic antidepressants such as protriptyline or clomipramine are extremely effective for cataplexy, but are rarely used due to their adverse effects.^2,47

FUTURE WORK

While our understanding of narcolepsy has advanced, there are still gaps in our knowledge of the disorder—namely, the specific trigger for the loss of hypocretin neurons in type 1 narcolepsy and the underlying pathophysiology of type 2.

A number of emerging therapies target the hypocretin system, including peptide replacement, neuronal transplant, and immunotherapy preventing hypocretin neuronal cell death.^50,54,55 Additional drugs designed to improve alertness that do not involve the hypocretin system are also being developed, including a histamine inverse agonist.^50,56 Sodium oxybate and modafinil, although currently approved for use in adults, are still off-label in pediatric practice. Studies of the safety and efficacy of these medications in children are needed.^7,57

But, after the results of many recent studies, it turns out I should not have been so comfortable after all. This should not be a surprise. We should never be overly confident with our understanding of anything in clinical practice.

In this issue, Dr. Martin Blaser discusses his work, which supports the hypothesis that the currently increased prevalence of obesity and diabetes is at least in part due to reduced diversity in the gut microbiome. The increased exposure to antibiotics through prescriptions for women before and during pregnancy, as well as perhaps their exposure to antibiotics in the environment, results in changes to the gut and vaginal flora that influence the developing gut and likely other anatomic microbiomes in the neonate and infant. Fascinating research done in mice, utilizing fecal transfer experiments, is building an evidence trail to support the concept that the microbiome plays a major role in the development of childhood and adult obesity, and the gut microbiome is influenced by its exposure to antibiotics, perhaps given years earlier.

Knowledge of the gastrointestinal and other human microbiomes is exploding. I now wonder how many seemingly random clinical events associated with antibiotic use that were not understood and were easily dismissed as stochastic warrant formal study. Some of my patients with rheumatoid arthritis have described flares after eating certain foods and transient remissions or exacerbations after treatment with antibiotics. An epidemiologic study has linked the likelihood of developing childhood inflammatory bowel disease with exposure to antibiotics. Even more fascinating are observations that the microbiota composition (influenced by antibiotics) can influence the outcome of cardiac allografts in a murine model and the response of certain tumors to immune checkpoint inhibitors in murine and human studies. The mechanism may relate to the effects of the microbiome on immune cell activation and migration. Several disorders have been linked to specific bacteria in the gut microbiome, and others as diverse as cardiovascular events and the acute inflammatory response to monosodium urate crystals (gout) are affected by metabolites generated by bacteria in the gut.

The use of germ-free and antibiotic-treated mice in the laboratory, with selective repopulation of their gut microbiome with flora harvested from other strains of mice or selected humans, will continue to teach us much about the role that these microbes and other inhabitants play in controlling normal and disease-disrupted homeostasis. C difficile overgrowth after antibiotic exposure, and the successful treatment of refractory C difficile with fecal transplantation,¹ was just the beginning.

The simple writing of a prescription for an antibiotic is a far more complicated and long-lasting affair than most of us have thought.

But, after the results of many recent studies, it turns out I should not have been so comfortable after all. This should not be a surprise. We should never be overly confident with our understanding of anything in clinical practice.

In this issue, Dr. Martin Blaser discusses his work, which supports the hypothesis that the currently increased prevalence of obesity and diabetes is at least in part due to reduced diversity in the gut microbiome. The increased exposure to antibiotics through prescriptions for women before and during pregnancy, as well as perhaps their exposure to antibiotics in the environment, results in changes to the gut and vaginal flora that influence the developing gut and likely other anatomic microbiomes in the neonate and infant. Fascinating research done in mice, utilizing fecal transfer experiments, is building an evidence trail to support the concept that the microbiome plays a major role in the development of childhood and adult obesity, and the gut microbiome is influenced by its exposure to antibiotics, perhaps given years earlier.

Knowledge of the gastrointestinal and other human microbiomes is exploding. I now wonder how many seemingly random clinical events associated with antibiotic use that were not understood and were easily dismissed as stochastic warrant formal study. Some of my patients with rheumatoid arthritis have described flares after eating certain foods and transient remissions or exacerbations after treatment with antibiotics. An epidemiologic study has linked the likelihood of developing childhood inflammatory bowel disease with exposure to antibiotics. Even more fascinating are observations that the microbiota composition (influenced by antibiotics) can influence the outcome of cardiac allografts in a murine model and the response of certain tumors to immune checkpoint inhibitors in murine and human studies. The mechanism may relate to the effects of the microbiome on immune cell activation and migration. Several disorders have been linked to specific bacteria in the gut microbiome, and others as diverse as cardiovascular events and the acute inflammatory response to monosodium urate crystals (gout) are affected by metabolites generated by bacteria in the gut.

The use of germ-free and antibiotic-treated mice in the laboratory, with selective repopulation of their gut microbiome with flora harvested from other strains of mice or selected humans, will continue to teach us much about the role that these microbes and other inhabitants play in controlling normal and disease-disrupted homeostasis. C difficile overgrowth after antibiotic exposure, and the successful treatment of refractory C difficile with fecal transplantation,¹ was just the beginning.

The simple writing of a prescription for an antibiotic is a far more complicated and long-lasting affair than most of us have thought.

But, after the results of many recent studies, it turns out I should not have been so comfortable after all. This should not be a surprise. We should never be overly confident with our understanding of anything in clinical practice.

In this issue, Dr. Martin Blaser discusses his work, which supports the hypothesis that the currently increased prevalence of obesity and diabetes is at least in part due to reduced diversity in the gut microbiome. The increased exposure to antibiotics through prescriptions for women before and during pregnancy, as well as perhaps their exposure to antibiotics in the environment, results in changes to the gut and vaginal flora that influence the developing gut and likely other anatomic microbiomes in the neonate and infant. Fascinating research done in mice, utilizing fecal transfer experiments, is building an evidence trail to support the concept that the microbiome plays a major role in the development of childhood and adult obesity, and the gut microbiome is influenced by its exposure to antibiotics, perhaps given years earlier.

Knowledge of the gastrointestinal and other human microbiomes is exploding. I now wonder how many seemingly random clinical events associated with antibiotic use that were not understood and were easily dismissed as stochastic warrant formal study. Some of my patients with rheumatoid arthritis have described flares after eating certain foods and transient remissions or exacerbations after treatment with antibiotics. An epidemiologic study has linked the likelihood of developing childhood inflammatory bowel disease with exposure to antibiotics. Even more fascinating are observations that the microbiota composition (influenced by antibiotics) can influence the outcome of cardiac allografts in a murine model and the response of certain tumors to immune checkpoint inhibitors in murine and human studies. The mechanism may relate to the effects of the microbiome on immune cell activation and migration. Several disorders have been linked to specific bacteria in the gut microbiome, and others as diverse as cardiovascular events and the acute inflammatory response to monosodium urate crystals (gout) are affected by metabolites generated by bacteria in the gut.

The use of germ-free and antibiotic-treated mice in the laboratory, with selective repopulation of their gut microbiome with flora harvested from other strains of mice or selected humans, will continue to teach us much about the role that these microbes and other inhabitants play in controlling normal and disease-disrupted homeostasis. C difficile overgrowth after antibiotic exposure, and the successful treatment of refractory C difficile with fecal transplantation,¹ was just the beginning.

The simple writing of a prescription for an antibiotic is a far more complicated and long-lasting affair than most of us have thought.

A 39-year-old woman presented to the emergency department with a 2-day history of exertional dyspnea, left-sided chest pain with pleuritic characteristics, and cough without fever or chills. She had a history of severe postprandial nausea and vomiting, weight loss, and malnutrition, which had necessitated placement of a peripherally inserted central catheter in her right arm for total parenteral nutrition.

On physical examination, the patient was afebrile but tachycardic and tachypneic. Her oxygen saturation on room air by pulse oximetry was 91%, though she was not in significant distress. Breath sounds were equal bilaterally and clear, with symmetrical chest wall expansion.

Her white blood cell count was 18.5 × 10⁹/L (reference range 3.5–10.5), with 19.3% eosinophils (reference range 1%–7%); her D-dimer level was also elevated.

Conditions to consider in a patient with these imaging findings in the setting of leukocytosis and eosinophilia include mycobacterial infection, hypersensitivity reaction, diffuse fungal infiltrates, and possibly metastatic disease such as thyroid carcinoma or melanoma. The patient reported having had a purified protein derivative test that was positive for tuberculosis, but she denied having had active disease.

She underwent bronchosocopy. Bronchoalveolar lavage specimen study showed an elevated eosinophil count of 17%. Acid-fast staining detected no organisms. Transbronchial biopsy study revealed foreign-body granulomas from microcrystalline cellulose microemboli deposited in the microvasculature of the patient’s lungs. Upon further questioning the patient admitted she had crushed oral tablets of prescribed opioids and injected them intravenously.

A COMPLICATION OF INJECTING ORAL TABLETS

Oral tablets typically contain talc, cellulose, cornstarch, or combinations of these substances as binding agents. When pulverized, the powder can be combined with water to form an injectable solution with higher and more rapid bioavailability.^1,2 The binders, however, are insoluble and accumulate in various tissues.

Intravenous injection of microcellulose has been shown to produce pulmonary and peripheral eosinophilia in birds. In humans, the immune response in foreign body granulomatosis can vary, and case reports have not mentioned eosinophils in the lungs or serum.^3,4

Deposition of these particles in pulmonary vessels is common and can trigger a potentially fatal reaction, presenting as acute onset of cough, chest pain, dyspnea, fever, and pulmonary hypertension. The severity of these clinical findings is relative to the degree of pulmonary hypertension created by the arteriolar involvement of these emboli.^2,5

Our patient’s exertional dyspnea and hypoxemia resolved during 1 week of hospitalization with conservative management and supplemental oxygen. She was referred to our pain rehabilitation clinic, where she was successfully weaned from narcotics. Her pulmonary findings on computed tomography were still present 3 years after her initial images, though less prominent.

A 39-year-old woman presented to the emergency department with a 2-day history of exertional dyspnea, left-sided chest pain with pleuritic characteristics, and cough without fever or chills. She had a history of severe postprandial nausea and vomiting, weight loss, and malnutrition, which had necessitated placement of a peripherally inserted central catheter in her right arm for total parenteral nutrition.

On physical examination, the patient was afebrile but tachycardic and tachypneic. Her oxygen saturation on room air by pulse oximetry was 91%, though she was not in significant distress. Breath sounds were equal bilaterally and clear, with symmetrical chest wall expansion.

Her white blood cell count was 18.5 × 10⁹/L (reference range 3.5–10.5), with 19.3% eosinophils (reference range 1%–7%); her D-dimer level was also elevated.

Conditions to consider in a patient with these imaging findings in the setting of leukocytosis and eosinophilia include mycobacterial infection, hypersensitivity reaction, diffuse fungal infiltrates, and possibly metastatic disease such as thyroid carcinoma or melanoma. The patient reported having had a purified protein derivative test that was positive for tuberculosis, but she denied having had active disease.

She underwent bronchosocopy. Bronchoalveolar lavage specimen study showed an elevated eosinophil count of 17%. Acid-fast staining detected no organisms. Transbronchial biopsy study revealed foreign-body granulomas from microcrystalline cellulose microemboli deposited in the microvasculature of the patient’s lungs. Upon further questioning the patient admitted she had crushed oral tablets of prescribed opioids and injected them intravenously.

A COMPLICATION OF INJECTING ORAL TABLETS

Oral tablets typically contain talc, cellulose, cornstarch, or combinations of these substances as binding agents. When pulverized, the powder can be combined with water to form an injectable solution with higher and more rapid bioavailability.^1,2 The binders, however, are insoluble and accumulate in various tissues.

Intravenous injection of microcellulose has been shown to produce pulmonary and peripheral eosinophilia in birds. In humans, the immune response in foreign body granulomatosis can vary, and case reports have not mentioned eosinophils in the lungs or serum.^3,4

Deposition of these particles in pulmonary vessels is common and can trigger a potentially fatal reaction, presenting as acute onset of cough, chest pain, dyspnea, fever, and pulmonary hypertension. The severity of these clinical findings is relative to the degree of pulmonary hypertension created by the arteriolar involvement of these emboli.^2,5

Our patient’s exertional dyspnea and hypoxemia resolved during 1 week of hospitalization with conservative management and supplemental oxygen. She was referred to our pain rehabilitation clinic, where she was successfully weaned from narcotics. Her pulmonary findings on computed tomography were still present 3 years after her initial images, though less prominent.

A 39-year-old woman presented to the emergency department with a 2-day history of exertional dyspnea, left-sided chest pain with pleuritic characteristics, and cough without fever or chills. She had a history of severe postprandial nausea and vomiting, weight loss, and malnutrition, which had necessitated placement of a peripherally inserted central catheter in her right arm for total parenteral nutrition.

On physical examination, the patient was afebrile but tachycardic and tachypneic. Her oxygen saturation on room air by pulse oximetry was 91%, though she was not in significant distress. Breath sounds were equal bilaterally and clear, with symmetrical chest wall expansion.

Her white blood cell count was 18.5 × 10⁹/L (reference range 3.5–10.5), with 19.3% eosinophils (reference range 1%–7%); her D-dimer level was also elevated.

Conditions to consider in a patient with these imaging findings in the setting of leukocytosis and eosinophilia include mycobacterial infection, hypersensitivity reaction, diffuse fungal infiltrates, and possibly metastatic disease such as thyroid carcinoma or melanoma. The patient reported having had a purified protein derivative test that was positive for tuberculosis, but she denied having had active disease.

She underwent bronchosocopy. Bronchoalveolar lavage specimen study showed an elevated eosinophil count of 17%. Acid-fast staining detected no organisms. Transbronchial biopsy study revealed foreign-body granulomas from microcrystalline cellulose microemboli deposited in the microvasculature of the patient’s lungs. Upon further questioning the patient admitted she had crushed oral tablets of prescribed opioids and injected them intravenously.

A COMPLICATION OF INJECTING ORAL TABLETS

Oral tablets typically contain talc, cellulose, cornstarch, or combinations of these substances as binding agents. When pulverized, the powder can be combined with water to form an injectable solution with higher and more rapid bioavailability.^1,2 The binders, however, are insoluble and accumulate in various tissues.

Intravenous injection of microcellulose has been shown to produce pulmonary and peripheral eosinophilia in birds. In humans, the immune response in foreign body granulomatosis can vary, and case reports have not mentioned eosinophils in the lungs or serum.^3,4

Deposition of these particles in pulmonary vessels is common and can trigger a potentially fatal reaction, presenting as acute onset of cough, chest pain, dyspnea, fever, and pulmonary hypertension. The severity of these clinical findings is relative to the degree of pulmonary hypertension created by the arteriolar involvement of these emboli.^2,5

Our patient’s exertional dyspnea and hypoxemia resolved during 1 week of hospitalization with conservative management and supplemental oxygen. She was referred to our pain rehabilitation clinic, where she was successfully weaned from narcotics. Her pulmonary findings on computed tomography were still present 3 years after her initial images, though less prominent.

An 82-year-old man with poorly controlled diabetes mellitus presented to our emergency department with a 1-day history of confusion and coffee-ground emesis.

Biopsy study revealed necrosis of the esophageal mucosa. A diagnosis of acute necrotizing esophagitis was made.

ACUTE NECROTIZING ESOPHAGITIS

Acute necrotizing esophagitis is thought to arise from a combination of an ischemic insult to the esophagus, an impaired mucosal barrier system, and a backflow injury from chemical contents of gastric secretions.¹ The tissue hypoperfusion derives from vasculopathy, hypotension, or malnutrition. It is associated with diabetes mellitus, diabetic ketoacidosis, lactic acidosis, alcohol abuse, cirrhosis, renal insufficiency, malignancy, antibiotic use, esophageal infections, and aortic dissection.

The esophagus has a diverse blood supply. The upper esophagus is supplied by the inferior thyroid arteries, the mid-esophagus by the bronchial, proper esophageal, and intercostal arteries, and the distal esophagus by the left gastric and left inferior phrenic arteries.¹

KEY FEATURES AND DIAGNOSTIC CLUES

The necrotic changes are prominent in the distal esophagus, which is more susceptible to ischemia and mucosal injury. The characteristic endoscopic finding is a diffuse black esophagus with a sharp transition to normal mucosa at the gastroesophageal junction.

The differential diagnosis includes melanosis, pseudomelanosis, malignant melanoma, acanthosis nigricans, coal dust deposition, caustic ingestion, radiation esophagitis, and infectious esophagitis caused by cytomegalovirus, herpes simplex virus, Candida albicans, or Klebsiella pneumoniae.^2–4

TREATMENT AND OUTCOME

Avoidance of oral intake and gastric acid suppression with intravenous proton pump inhibitors are recommended to prevent additional injury of the esophageal mucosa.

The condition generally resolves with restored blood flow and treatment of any coexisting illness. However, it may be complicated by perforation (6.8%), mediastinitis (5.7%), or subsequent development of esophageal stricture (10.2%).⁵ Patients with esophageal stricture require endoscopic dilation after mucosal recovery.

The overall risk of death in acute necrotizing esophagitis is high (31.8%) and most often due to the underlying disease, such as sepsis, malignancy, cardiogenic shock, or hypovolemic shock.⁵ The mortality rate directly attributed to complications of acute necrotizing esophagitis is much lower (5.7%).⁵     

An 82-year-old man with poorly controlled diabetes mellitus presented to our emergency department with a 1-day history of confusion and coffee-ground emesis.

Biopsy study revealed necrosis of the esophageal mucosa. A diagnosis of acute necrotizing esophagitis was made.

ACUTE NECROTIZING ESOPHAGITIS

Acute necrotizing esophagitis is thought to arise from a combination of an ischemic insult to the esophagus, an impaired mucosal barrier system, and a backflow injury from chemical contents of gastric secretions.¹ The tissue hypoperfusion derives from vasculopathy, hypotension, or malnutrition. It is associated with diabetes mellitus, diabetic ketoacidosis, lactic acidosis, alcohol abuse, cirrhosis, renal insufficiency, malignancy, antibiotic use, esophageal infections, and aortic dissection.

The esophagus has a diverse blood supply. The upper esophagus is supplied by the inferior thyroid arteries, the mid-esophagus by the bronchial, proper esophageal, and intercostal arteries, and the distal esophagus by the left gastric and left inferior phrenic arteries.¹

KEY FEATURES AND DIAGNOSTIC CLUES

The necrotic changes are prominent in the distal esophagus, which is more susceptible to ischemia and mucosal injury. The characteristic endoscopic finding is a diffuse black esophagus with a sharp transition to normal mucosa at the gastroesophageal junction.

The differential diagnosis includes melanosis, pseudomelanosis, malignant melanoma, acanthosis nigricans, coal dust deposition, caustic ingestion, radiation esophagitis, and infectious esophagitis caused by cytomegalovirus, herpes simplex virus, Candida albicans, or Klebsiella pneumoniae.^2–4

TREATMENT AND OUTCOME

Avoidance of oral intake and gastric acid suppression with intravenous proton pump inhibitors are recommended to prevent additional injury of the esophageal mucosa.

The condition generally resolves with restored blood flow and treatment of any coexisting illness. However, it may be complicated by perforation (6.8%), mediastinitis (5.7%), or subsequent development of esophageal stricture (10.2%).⁵ Patients with esophageal stricture require endoscopic dilation after mucosal recovery.

The overall risk of death in acute necrotizing esophagitis is high (31.8%) and most often due to the underlying disease, such as sepsis, malignancy, cardiogenic shock, or hypovolemic shock.⁵ The mortality rate directly attributed to complications of acute necrotizing esophagitis is much lower (5.7%).⁵     

An 82-year-old man with poorly controlled diabetes mellitus presented to our emergency department with a 1-day history of confusion and coffee-ground emesis.

Biopsy study revealed necrosis of the esophageal mucosa. A diagnosis of acute necrotizing esophagitis was made.

ACUTE NECROTIZING ESOPHAGITIS

Acute necrotizing esophagitis is thought to arise from a combination of an ischemic insult to the esophagus, an impaired mucosal barrier system, and a backflow injury from chemical contents of gastric secretions.¹ The tissue hypoperfusion derives from vasculopathy, hypotension, or malnutrition. It is associated with diabetes mellitus, diabetic ketoacidosis, lactic acidosis, alcohol abuse, cirrhosis, renal insufficiency, malignancy, antibiotic use, esophageal infections, and aortic dissection.

The esophagus has a diverse blood supply. The upper esophagus is supplied by the inferior thyroid arteries, the mid-esophagus by the bronchial, proper esophageal, and intercostal arteries, and the distal esophagus by the left gastric and left inferior phrenic arteries.¹

KEY FEATURES AND DIAGNOSTIC CLUES

The necrotic changes are prominent in the distal esophagus, which is more susceptible to ischemia and mucosal injury. The characteristic endoscopic finding is a diffuse black esophagus with a sharp transition to normal mucosa at the gastroesophageal junction.

The differential diagnosis includes melanosis, pseudomelanosis, malignant melanoma, acanthosis nigricans, coal dust deposition, caustic ingestion, radiation esophagitis, and infectious esophagitis caused by cytomegalovirus, herpes simplex virus, Candida albicans, or Klebsiella pneumoniae.^2–4

TREATMENT AND OUTCOME

Avoidance of oral intake and gastric acid suppression with intravenous proton pump inhibitors are recommended to prevent additional injury of the esophageal mucosa.

The condition generally resolves with restored blood flow and treatment of any coexisting illness. However, it may be complicated by perforation (6.8%), mediastinitis (5.7%), or subsequent development of esophageal stricture (10.2%).⁵ Patients with esophageal stricture require endoscopic dilation after mucosal recovery.

The overall risk of death in acute necrotizing esophagitis is high (31.8%) and most often due to the underlying disease, such as sepsis, malignancy, cardiogenic shock, or hypovolemic shock.⁵ The mortality rate directly attributed to complications of acute necrotizing esophagitis is much lower (5.7%).⁵     

Vasovagal syncope is usually benign, and although it often recurs, increasing fluid and salt intake and performing counter-pressure maneuvers are usually sufficient.¹ However, if patients continue to have syncopal episodes despite these first-line measures, other options include drug therapy with midodrine, fludrocortisone, beta-blockers, or selective serotonin reuptake inhibitors; orthostatic training; and, in some cases, pacemaker implantation. The 2017 guidelines from the American College of Cardiology, American Heart Association, and Heart Rhythm Society (ACC/AHA/HRS) are helpful in the management of these patients.¹

RATIONALE

Although vasovagal syncope is considered benign, it can result in injury and can significantly affect quality of life.

The diagnosis can often be established in the initial evaluation with a structured history, physical examination, and electrocardiography. If the diagnosis is still unclear, tilt-table testing can be useful and has an ACC/AHA/HRS class IIa (moderate) recommendation.¹ Once the diagnosis of vasovagal syncope is made, first-line measures can be instituted.

FIRST-LINE MEASURES

An explanation of the diagnosis, education on avoiding triggers such as prolonged standing and warm environments, coping with potentially stressful visits to the doctor or dentist, and reassurance that the condition is benign are all strongly recommended (class I).¹

Initial measures include performing physical counter-pressure maneuvers (class IIa), increasing salt and fluid intake (class IIb) in the absence of contraindications, and, in selected patients, reducing or withdrawing hypotensive medications when appropriate (class IIb).

Physical counter-pressure maneuvers are recommended for patients whose syncopal episodes have a sufficiently long prodromal period. Maneuvers include the following:

• Leg crossing: crossing the legs while tensing leg, abdominal, and buttock muscles
• Handgrip: maximally contracting a rubber ball or other object in the dominant hand
• Squatting
• Limb or abdominal contractions
• Arm tensing: contracting both arms by gripping one hand with the other and abducting both arms.²

The effectiveness of counter-pressure maneuvers was studied by van Dijk et al² in a multicenter prospective randomized clinical trial that included 223 patients with recurrent vasovagal syncope associated with prodromal symptoms. They concluded that these maneuvers decreased the recurrence of syncopal episodes, with a relative risk reduction of 0.36 (95% confidence interval 0.11–0.53, P < .005) and were low-cost and risk-free.

Confirming the diagnosis of vasovagal syncope with tilt-table testing may reassure the patient. It can also help the patient learn to identify the symptoms associated with a vasovagal episode, which in turn may encourage timely use of physical counter-pressure maneuvers at the onset.

The evidence for increasing salt and fluid intake for patients with vasovagal syncope is limited. But in the absence of a contraindication such as hypertension, renal disease, or heart failure, it may be reasonable to encourage the ingestion of 2 L to 3 L of fluid per day and a total of 6 g to 9 g of salt per day (around 1 to 2 heaping teaspoons of salt).¹

 

 

MEDICAL THERAPY

In patients who continue to have syncopal episodes despite adequate use of first-line measures, medical therapy can be considered. Unfortunately, evidence supporting drug therapy for recurrent syncope is limited.³ Options include midodrine (class IIa), fludrocortisone (class IIb), beta-blockers (class IIb), and selective serotonin reuptake inhibitors (class IIb).¹

Midodrine has the strongest recommendation and is a reasonable option if there is no history of hypertension, heart failure, or urinary retention. It is a peripheral alpha-agonist that ameliorates the reduction in peripheral sympathetic neural outflow responsible for venous pooling and vasodepression in vasovagal syncope.^4–6

Fludrocortisone results in increased blood volume due to mineralocorticoid activity. In the Prevention of Syncope Trial 2 of fludrocortisone vs placebo, patients on fludrocortisone had a “marginally nonsignificant” reduction in recurrence of syncope over 1 year (hazard ratio 0.69, P = .069).⁷

Overall, beta-blockers have failed to prevent syncope in randomized controlled trials. But in a meta-analysis that included patients from the Prevention of Syncope Trial,⁸ an age-dependent benefit of beta-blockers was noted in patients age 42 and older.⁹ Therefore, a beta-blocker may be a reasonable option in patients in this age group with recurrent vasovagal syncope.¹

Serotonin has central effects on blood pressure and heart rate that can induce syncope. However, evidence for the effectiveness of selective serotonin reuptake inhibitors in the prevention of recurrent vasovagal syncope has been contradictory in small trials.^10,11

When choosing a drug, contraindications should be considered, including possible effects during pregnancy in women of childbearing age.

OTHER MEASURES

Orthostatic training, with repetitive tilt-table testing until a test is negative, or with daily standing quietly against a wall for prolonged periods of time, has not been shown to have sustained benefit in reducing the recurrence of syncopal episodes (class IIb recommendation).¹

Dual-chamber pacing can be considered in carefully selected patients age 40 or older with syncope and documented asystole of at least 3 seconds or spontaneous pauses of at least 6 seconds without syncope on implantable loop recorder monitoring (class IIb recommendation).^1,12,13 Strict patient selection increases the likelihood that pacing will be effective.¹ For example, patients with documented asystole during syncope and a tilt-table test that induces minimal or no vasodepressor response are more likely to respond than patients with a positive tilt-table test with a vasodepressor (hypotensive) response.¹³

Tilt-table testing may also be considered to identify patients with a hypotensive response who would be less likely to respond to permanent cardiac pacing.¹⁴ 


Compression garments carry a class IIa recommendation for orthostatic hypotension,1 but they have not been adequately studied in vasovagal syncope.

Vasovagal syncope is usually benign, and although it often recurs, increasing fluid and salt intake and performing counter-pressure maneuvers are usually sufficient.¹ However, if patients continue to have syncopal episodes despite these first-line measures, other options include drug therapy with midodrine, fludrocortisone, beta-blockers, or selective serotonin reuptake inhibitors; orthostatic training; and, in some cases, pacemaker implantation. The 2017 guidelines from the American College of Cardiology, American Heart Association, and Heart Rhythm Society (ACC/AHA/HRS) are helpful in the management of these patients.¹

RATIONALE

Although vasovagal syncope is considered benign, it can result in injury and can significantly affect quality of life.

The diagnosis can often be established in the initial evaluation with a structured history, physical examination, and electrocardiography. If the diagnosis is still unclear, tilt-table testing can be useful and has an ACC/AHA/HRS class IIa (moderate) recommendation.¹ Once the diagnosis of vasovagal syncope is made, first-line measures can be instituted.

FIRST-LINE MEASURES

An explanation of the diagnosis, education on avoiding triggers such as prolonged standing and warm environments, coping with potentially stressful visits to the doctor or dentist, and reassurance that the condition is benign are all strongly recommended (class I).¹

Initial measures include performing physical counter-pressure maneuvers (class IIa), increasing salt and fluid intake (class IIb) in the absence of contraindications, and, in selected patients, reducing or withdrawing hypotensive medications when appropriate (class IIb).

Physical counter-pressure maneuvers are recommended for patients whose syncopal episodes have a sufficiently long prodromal period. Maneuvers include the following:

• Leg crossing: crossing the legs while tensing leg, abdominal, and buttock muscles
• Handgrip: maximally contracting a rubber ball or other object in the dominant hand
• Squatting
• Limb or abdominal contractions
• Arm tensing: contracting both arms by gripping one hand with the other and abducting both arms.²

The effectiveness of counter-pressure maneuvers was studied by van Dijk et al² in a multicenter prospective randomized clinical trial that included 223 patients with recurrent vasovagal syncope associated with prodromal symptoms. They concluded that these maneuvers decreased the recurrence of syncopal episodes, with a relative risk reduction of 0.36 (95% confidence interval 0.11–0.53, P < .005) and were low-cost and risk-free.

Confirming the diagnosis of vasovagal syncope with tilt-table testing may reassure the patient. It can also help the patient learn to identify the symptoms associated with a vasovagal episode, which in turn may encourage timely use of physical counter-pressure maneuvers at the onset.

The evidence for increasing salt and fluid intake for patients with vasovagal syncope is limited. But in the absence of a contraindication such as hypertension, renal disease, or heart failure, it may be reasonable to encourage the ingestion of 2 L to 3 L of fluid per day and a total of 6 g to 9 g of salt per day (around 1 to 2 heaping teaspoons of salt).¹

 

 

MEDICAL THERAPY

In patients who continue to have syncopal episodes despite adequate use of first-line measures, medical therapy can be considered. Unfortunately, evidence supporting drug therapy for recurrent syncope is limited.³ Options include midodrine (class IIa), fludrocortisone (class IIb), beta-blockers (class IIb), and selective serotonin reuptake inhibitors (class IIb).¹

Midodrine has the strongest recommendation and is a reasonable option if there is no history of hypertension, heart failure, or urinary retention. It is a peripheral alpha-agonist that ameliorates the reduction in peripheral sympathetic neural outflow responsible for venous pooling and vasodepression in vasovagal syncope.^4–6

Fludrocortisone results in increased blood volume due to mineralocorticoid activity. In the Prevention of Syncope Trial 2 of fludrocortisone vs placebo, patients on fludrocortisone had a “marginally nonsignificant” reduction in recurrence of syncope over 1 year (hazard ratio 0.69, P = .069).⁷

Overall, beta-blockers have failed to prevent syncope in randomized controlled trials. But in a meta-analysis that included patients from the Prevention of Syncope Trial,⁸ an age-dependent benefit of beta-blockers was noted in patients age 42 and older.⁹ Therefore, a beta-blocker may be a reasonable option in patients in this age group with recurrent vasovagal syncope.¹

Serotonin has central effects on blood pressure and heart rate that can induce syncope. However, evidence for the effectiveness of selective serotonin reuptake inhibitors in the prevention of recurrent vasovagal syncope has been contradictory in small trials.^10,11

When choosing a drug, contraindications should be considered, including possible effects during pregnancy in women of childbearing age.

OTHER MEASURES

Orthostatic training, with repetitive tilt-table testing until a test is negative, or with daily standing quietly against a wall for prolonged periods of time, has not been shown to have sustained benefit in reducing the recurrence of syncopal episodes (class IIb recommendation).¹

Dual-chamber pacing can be considered in carefully selected patients age 40 or older with syncope and documented asystole of at least 3 seconds or spontaneous pauses of at least 6 seconds without syncope on implantable loop recorder monitoring (class IIb recommendation).^1,12,13 Strict patient selection increases the likelihood that pacing will be effective.¹ For example, patients with documented asystole during syncope and a tilt-table test that induces minimal or no vasodepressor response are more likely to respond than patients with a positive tilt-table test with a vasodepressor (hypotensive) response.¹³

Tilt-table testing may also be considered to identify patients with a hypotensive response who would be less likely to respond to permanent cardiac pacing.¹⁴ 


Compression garments carry a class IIa recommendation for orthostatic hypotension,1 but they have not been adequately studied in vasovagal syncope.

Vasovagal syncope is usually benign, and although it often recurs, increasing fluid and salt intake and performing counter-pressure maneuvers are usually sufficient.¹ However, if patients continue to have syncopal episodes despite these first-line measures, other options include drug therapy with midodrine, fludrocortisone, beta-blockers, or selective serotonin reuptake inhibitors; orthostatic training; and, in some cases, pacemaker implantation. The 2017 guidelines from the American College of Cardiology, American Heart Association, and Heart Rhythm Society (ACC/AHA/HRS) are helpful in the management of these patients.¹

RATIONALE

Although vasovagal syncope is considered benign, it can result in injury and can significantly affect quality of life.

The diagnosis can often be established in the initial evaluation with a structured history, physical examination, and electrocardiography. If the diagnosis is still unclear, tilt-table testing can be useful and has an ACC/AHA/HRS class IIa (moderate) recommendation.¹ Once the diagnosis of vasovagal syncope is made, first-line measures can be instituted.

FIRST-LINE MEASURES

An explanation of the diagnosis, education on avoiding triggers such as prolonged standing and warm environments, coping with potentially stressful visits to the doctor or dentist, and reassurance that the condition is benign are all strongly recommended (class I).¹

Initial measures include performing physical counter-pressure maneuvers (class IIa), increasing salt and fluid intake (class IIb) in the absence of contraindications, and, in selected patients, reducing or withdrawing hypotensive medications when appropriate (class IIb).

Physical counter-pressure maneuvers are recommended for patients whose syncopal episodes have a sufficiently long prodromal period. Maneuvers include the following:

• Leg crossing: crossing the legs while tensing leg, abdominal, and buttock muscles
• Handgrip: maximally contracting a rubber ball or other object in the dominant hand
• Squatting
• Limb or abdominal contractions
• Arm tensing: contracting both arms by gripping one hand with the other and abducting both arms.²

The effectiveness of counter-pressure maneuvers was studied by van Dijk et al² in a multicenter prospective randomized clinical trial that included 223 patients with recurrent vasovagal syncope associated with prodromal symptoms. They concluded that these maneuvers decreased the recurrence of syncopal episodes, with a relative risk reduction of 0.36 (95% confidence interval 0.11–0.53, P < .005) and were low-cost and risk-free.

Confirming the diagnosis of vasovagal syncope with tilt-table testing may reassure the patient. It can also help the patient learn to identify the symptoms associated with a vasovagal episode, which in turn may encourage timely use of physical counter-pressure maneuvers at the onset.

The evidence for increasing salt and fluid intake for patients with vasovagal syncope is limited. But in the absence of a contraindication such as hypertension, renal disease, or heart failure, it may be reasonable to encourage the ingestion of 2 L to 3 L of fluid per day and a total of 6 g to 9 g of salt per day (around 1 to 2 heaping teaspoons of salt).¹

 

 

MEDICAL THERAPY

In patients who continue to have syncopal episodes despite adequate use of first-line measures, medical therapy can be considered. Unfortunately, evidence supporting drug therapy for recurrent syncope is limited.³ Options include midodrine (class IIa), fludrocortisone (class IIb), beta-blockers (class IIb), and selective serotonin reuptake inhibitors (class IIb).¹

Midodrine has the strongest recommendation and is a reasonable option if there is no history of hypertension, heart failure, or urinary retention. It is a peripheral alpha-agonist that ameliorates the reduction in peripheral sympathetic neural outflow responsible for venous pooling and vasodepression in vasovagal syncope.^4–6

Fludrocortisone results in increased blood volume due to mineralocorticoid activity. In the Prevention of Syncope Trial 2 of fludrocortisone vs placebo, patients on fludrocortisone had a “marginally nonsignificant” reduction in recurrence of syncope over 1 year (hazard ratio 0.69, P = .069).⁷

Overall, beta-blockers have failed to prevent syncope in randomized controlled trials. But in a meta-analysis that included patients from the Prevention of Syncope Trial,⁸ an age-dependent benefit of beta-blockers was noted in patients age 42 and older.⁹ Therefore, a beta-blocker may be a reasonable option in patients in this age group with recurrent vasovagal syncope.¹

Serotonin has central effects on blood pressure and heart rate that can induce syncope. However, evidence for the effectiveness of selective serotonin reuptake inhibitors in the prevention of recurrent vasovagal syncope has been contradictory in small trials.^10,11

When choosing a drug, contraindications should be considered, including possible effects during pregnancy in women of childbearing age.

OTHER MEASURES

Orthostatic training, with repetitive tilt-table testing until a test is negative, or with daily standing quietly against a wall for prolonged periods of time, has not been shown to have sustained benefit in reducing the recurrence of syncopal episodes (class IIb recommendation).¹

Dual-chamber pacing can be considered in carefully selected patients age 40 or older with syncope and documented asystole of at least 3 seconds or spontaneous pauses of at least 6 seconds without syncope on implantable loop recorder monitoring (class IIb recommendation).^1,12,13 Strict patient selection increases the likelihood that pacing will be effective.¹ For example, patients with documented asystole during syncope and a tilt-table test that induces minimal or no vasodepressor response are more likely to respond than patients with a positive tilt-table test with a vasodepressor (hypotensive) response.¹³

Tilt-table testing may also be considered to identify patients with a hypotensive response who would be less likely to respond to permanent cardiac pacing.¹⁴ 


Compression garments carry a class IIa recommendation for orthostatic hypotension,1 but they have not been adequately studied in vasovagal syncope.

This final month of the year is the giving season, and many people delay their charitable contributions until then. Please consider making a difference in the world of vascular surgery this holiday season by donating to the SVS Foundation. The mission of the Foundation has expanded, and your dollars are needed now, more than ever. From research, to disaster relief efforts, to patient awareness and disease prevention, all contributions are ultimately aimed at improving patient health. Please give today.

This final month of the year is the giving season, and many people delay their charitable contributions until then. Please consider making a difference in the world of vascular surgery this holiday season by donating to the SVS Foundation. The mission of the Foundation has expanded, and your dollars are needed now, more than ever. From research, to disaster relief efforts, to patient awareness and disease prevention, all contributions are ultimately aimed at improving patient health. Please give today.

This final month of the year is the giving season, and many people delay their charitable contributions until then. Please consider making a difference in the world of vascular surgery this holiday season by donating to the SVS Foundation. The mission of the Foundation has expanded, and your dollars are needed now, more than ever. From research, to disaster relief efforts, to patient awareness and disease prevention, all contributions are ultimately aimed at improving patient health. Please give today.

Yes, anti-tumor necrosis factor (anti-TNF) therapy for inflammatory bowel disease (IBD) can be continued during pregnancy.

IBD is often diagnosed and treated in women during their reproductive years. Consequently, these patients face important decisions about the management of their disease and the safety of their baby. Clinicians should be prepared to offer guidance by discussing the risks and benefits of anti-TNF agents with their pregnant patients who have IBD, as well as with those considering pregnancy.

STUDIES OF THE POTENTIAL RISKS

Anti-TNF agents are monoclonal antibodies. Infliximab, adalimumab, and golimumab are actively transported into the fetal circulation via the placenta, mainly during the second and third trimesters. Certolizumab crosses the placenta only by passive means, because it lacks the fragment crystallizable (Fc) region required for placental transfer.¹

Effects on pregnancy outcomes

In a 2016 meta-analysis,² of 1,242 pregnancies in women with IBD, 482 were in women on anti-TNF therapy. It found no statistically significant difference in rates of adverse pregnancy outcomes including congenital abnormality, preterm birth, and low birth weight.

A meta-analysis of 1,216 pregnant women with IBD found no statistically significant differences in rates of spontaneous or elective abortion, preterm birth, low birth weight, or congenital malformation in those on anti-TNF therapy vs controls.³

A systematic review of 58 studies including more than 1,500 pregnant women with IBD who were exposed to anti-TNF agents concluded that there was no association with adverse pregnancy outcomes such as spontaneous abortion, preterm delivery, stillbirth, low birth weight, congenital malformation, or infection.⁴

A retrospective cohort study of 66 pregnant patients with IBD from several centers in Spain found that anti-TNF or thiopurine therapy during pregnancy did not increase the risk of pregnancy complications or neonatal complications.⁵

Effects on newborns

Cord blood studies have shown that maternal use of infliximab and adalimumab results in a detectable serum level in newborns, while cord blood levels of certolizumab are much lower.^1,6 In some studies, anti-TNF drugs were detectable in infants for up to 6 months after birth, whereas other studies found that detectable serum levels dropped soon after birth.^1,7

Addressing concern about an increased risk of infection or dysfunctional immune development in newborns exposed to anti-TNF drugs in utero, a systematic review found no increased risk.⁴ A retrospective multicenter cohort study of 841 children also reported no association between in utero exposure to anti-TNF agents and risk of severe infection in the short term or long term (mean of 4 years).⁸ Additional studies are under way to determine long-term risk to the newborn.⁷

 

 

THE TORONTO CONSENSUS GUIDELINES

The Toronto consensus guidelines strongly recommend continuing anti-TNF therapy during pregnancy in women with IBD who began maintenance therapy before conception.⁶

If a patient strongly prefers to stop therapy during pregnancy to limit fetal exposure, the Toronto consensus recommends giving the last dose at 22 to 24 weeks of gestation. However, this should only be considered in patients whose IBD is in remission and at low risk of relapse.^6,9

Although anti-TNF drugs may differ in terms of placental transfer, agents should not be switched in stable patients, as switching increases the risk of relapse.¹⁰

BENEFITS OF CONTINUING THERAPY

Active IBD poses a significantly greater risk to the mother and the baby than continuing anti-TNF therapy during pregnancy.^1,7 The primary benefit of continuing therapy is to maintain disease remission.

Among women with active IBD at the time of conception, one-third will have improvement in disease activity during the course of their pregnancy, one-third will have no change, and one-third will have worsening of disease activity. But if IBD is in remission at the time of conception, it will remain in remission in nearly 80% of women during pregnancy.¹

Women with active IBD are at increased risk of preterm delivery, low birth weight, and intrauterine growth restriction.^1,2,5 Also, women with IBD have an increased risk of venous thromboembolism, particularly if they have active disease during pregnancy.¹ Therefore, achieving and maintaining remission are vital in the management of the pregnant patient with IBD.

CONSIDERATIONS AFTER BIRTH: BREAST-FEEDING AND VACCINATION

Breast-feeding is considered safe. Minuscule amounts of infliximab or adalimumab are transferred in breast milk but are unlikely to result in systemic immune suppression in the infant.⁷

Live-attenuated vaccines should be avoided for the first 6 months in infants exposed to anti-TNF agents in utero.^1,7,11 All other vaccines, including hepatitis B virus vaccine, should be given according to standard schedules.⁶

OUR RECOMMENDATIONS

The goal of managing IBD in women of reproductive age is to minimize the risk of adverse outcomes for both mother and baby. We recommend a team approach, working closely with a gastroenterologist and a high-risk-pregnancy obstetrician, if available.

Patients should continue anti-TNF therapy during pregnancy because evidence supports its safety. If a woman wants to stop therapy and is at low risk of relapse, we recommend giving the last dose at 22 to 24 weeks of gestation, then promptly resuming therapy postpartum.

Live-attenuated vaccines (eg, influenza, rotavirus) should be avoided for the first 6 months in babies born to mothers on anti-TNF therapy.

Yes, anti-tumor necrosis factor (anti-TNF) therapy for inflammatory bowel disease (IBD) can be continued during pregnancy.

IBD is often diagnosed and treated in women during their reproductive years. Consequently, these patients face important decisions about the management of their disease and the safety of their baby. Clinicians should be prepared to offer guidance by discussing the risks and benefits of anti-TNF agents with their pregnant patients who have IBD, as well as with those considering pregnancy.

STUDIES OF THE POTENTIAL RISKS

Anti-TNF agents are monoclonal antibodies. Infliximab, adalimumab, and golimumab are actively transported into the fetal circulation via the placenta, mainly during the second and third trimesters. Certolizumab crosses the placenta only by passive means, because it lacks the fragment crystallizable (Fc) region required for placental transfer.¹

Effects on pregnancy outcomes

In a 2016 meta-analysis,² of 1,242 pregnancies in women with IBD, 482 were in women on anti-TNF therapy. It found no statistically significant difference in rates of adverse pregnancy outcomes including congenital abnormality, preterm birth, and low birth weight.

A meta-analysis of 1,216 pregnant women with IBD found no statistically significant differences in rates of spontaneous or elective abortion, preterm birth, low birth weight, or congenital malformation in those on anti-TNF therapy vs controls.³

A systematic review of 58 studies including more than 1,500 pregnant women with IBD who were exposed to anti-TNF agents concluded that there was no association with adverse pregnancy outcomes such as spontaneous abortion, preterm delivery, stillbirth, low birth weight, congenital malformation, or infection.⁴

A retrospective cohort study of 66 pregnant patients with IBD from several centers in Spain found that anti-TNF or thiopurine therapy during pregnancy did not increase the risk of pregnancy complications or neonatal complications.⁵

Effects on newborns

Cord blood studies have shown that maternal use of infliximab and adalimumab results in a detectable serum level in newborns, while cord blood levels of certolizumab are much lower.^1,6 In some studies, anti-TNF drugs were detectable in infants for up to 6 months after birth, whereas other studies found that detectable serum levels dropped soon after birth.^1,7

Addressing concern about an increased risk of infection or dysfunctional immune development in newborns exposed to anti-TNF drugs in utero, a systematic review found no increased risk.⁴ A retrospective multicenter cohort study of 841 children also reported no association between in utero exposure to anti-TNF agents and risk of severe infection in the short term or long term (mean of 4 years).⁸ Additional studies are under way to determine long-term risk to the newborn.⁷

 

 

THE TORONTO CONSENSUS GUIDELINES

The Toronto consensus guidelines strongly recommend continuing anti-TNF therapy during pregnancy in women with IBD who began maintenance therapy before conception.⁶

If a patient strongly prefers to stop therapy during pregnancy to limit fetal exposure, the Toronto consensus recommends giving the last dose at 22 to 24 weeks of gestation. However, this should only be considered in patients whose IBD is in remission and at low risk of relapse.^6,9

Although anti-TNF drugs may differ in terms of placental transfer, agents should not be switched in stable patients, as switching increases the risk of relapse.¹⁰

BENEFITS OF CONTINUING THERAPY

Active IBD poses a significantly greater risk to the mother and the baby than continuing anti-TNF therapy during pregnancy.^1,7 The primary benefit of continuing therapy is to maintain disease remission.

Among women with active IBD at the time of conception, one-third will have improvement in disease activity during the course of their pregnancy, one-third will have no change, and one-third will have worsening of disease activity. But if IBD is in remission at the time of conception, it will remain in remission in nearly 80% of women during pregnancy.¹

Women with active IBD are at increased risk of preterm delivery, low birth weight, and intrauterine growth restriction.^1,2,5 Also, women with IBD have an increased risk of venous thromboembolism, particularly if they have active disease during pregnancy.¹ Therefore, achieving and maintaining remission are vital in the management of the pregnant patient with IBD.

CONSIDERATIONS AFTER BIRTH: BREAST-FEEDING AND VACCINATION

Breast-feeding is considered safe. Minuscule amounts of infliximab or adalimumab are transferred in breast milk but are unlikely to result in systemic immune suppression in the infant.⁷

Live-attenuated vaccines should be avoided for the first 6 months in infants exposed to anti-TNF agents in utero.^1,7,11 All other vaccines, including hepatitis B virus vaccine, should be given according to standard schedules.⁶

OUR RECOMMENDATIONS

The goal of managing IBD in women of reproductive age is to minimize the risk of adverse outcomes for both mother and baby. We recommend a team approach, working closely with a gastroenterologist and a high-risk-pregnancy obstetrician, if available.

Patients should continue anti-TNF therapy during pregnancy because evidence supports its safety. If a woman wants to stop therapy and is at low risk of relapse, we recommend giving the last dose at 22 to 24 weeks of gestation, then promptly resuming therapy postpartum.

Live-attenuated vaccines (eg, influenza, rotavirus) should be avoided for the first 6 months in babies born to mothers on anti-TNF therapy.

Yes, anti-tumor necrosis factor (anti-TNF) therapy for inflammatory bowel disease (IBD) can be continued during pregnancy.

IBD is often diagnosed and treated in women during their reproductive years. Consequently, these patients face important decisions about the management of their disease and the safety of their baby. Clinicians should be prepared to offer guidance by discussing the risks and benefits of anti-TNF agents with their pregnant patients who have IBD, as well as with those considering pregnancy.

STUDIES OF THE POTENTIAL RISKS

Anti-TNF agents are monoclonal antibodies. Infliximab, adalimumab, and golimumab are actively transported into the fetal circulation via the placenta, mainly during the second and third trimesters. Certolizumab crosses the placenta only by passive means, because it lacks the fragment crystallizable (Fc) region required for placental transfer.¹

Effects on pregnancy outcomes

In a 2016 meta-analysis,² of 1,242 pregnancies in women with IBD, 482 were in women on anti-TNF therapy. It found no statistically significant difference in rates of adverse pregnancy outcomes including congenital abnormality, preterm birth, and low birth weight.

A meta-analysis of 1,216 pregnant women with IBD found no statistically significant differences in rates of spontaneous or elective abortion, preterm birth, low birth weight, or congenital malformation in those on anti-TNF therapy vs controls.³

A systematic review of 58 studies including more than 1,500 pregnant women with IBD who were exposed to anti-TNF agents concluded that there was no association with adverse pregnancy outcomes such as spontaneous abortion, preterm delivery, stillbirth, low birth weight, congenital malformation, or infection.⁴

A retrospective cohort study of 66 pregnant patients with IBD from several centers in Spain found that anti-TNF or thiopurine therapy during pregnancy did not increase the risk of pregnancy complications or neonatal complications.⁵

Effects on newborns

Cord blood studies have shown that maternal use of infliximab and adalimumab results in a detectable serum level in newborns, while cord blood levels of certolizumab are much lower.^1,6 In some studies, anti-TNF drugs were detectable in infants for up to 6 months after birth, whereas other studies found that detectable serum levels dropped soon after birth.^1,7

Addressing concern about an increased risk of infection or dysfunctional immune development in newborns exposed to anti-TNF drugs in utero, a systematic review found no increased risk.⁴ A retrospective multicenter cohort study of 841 children also reported no association between in utero exposure to anti-TNF agents and risk of severe infection in the short term or long term (mean of 4 years).⁸ Additional studies are under way to determine long-term risk to the newborn.⁷

 

 

THE TORONTO CONSENSUS GUIDELINES

The Toronto consensus guidelines strongly recommend continuing anti-TNF therapy during pregnancy in women with IBD who began maintenance therapy before conception.⁶

If a patient strongly prefers to stop therapy during pregnancy to limit fetal exposure, the Toronto consensus recommends giving the last dose at 22 to 24 weeks of gestation. However, this should only be considered in patients whose IBD is in remission and at low risk of relapse.^6,9

Although anti-TNF drugs may differ in terms of placental transfer, agents should not be switched in stable patients, as switching increases the risk of relapse.¹⁰

BENEFITS OF CONTINUING THERAPY

Active IBD poses a significantly greater risk to the mother and the baby than continuing anti-TNF therapy during pregnancy.^1,7 The primary benefit of continuing therapy is to maintain disease remission.

Among women with active IBD at the time of conception, one-third will have improvement in disease activity during the course of their pregnancy, one-third will have no change, and one-third will have worsening of disease activity. But if IBD is in remission at the time of conception, it will remain in remission in nearly 80% of women during pregnancy.¹

Women with active IBD are at increased risk of preterm delivery, low birth weight, and intrauterine growth restriction.^1,2,5 Also, women with IBD have an increased risk of venous thromboembolism, particularly if they have active disease during pregnancy.¹ Therefore, achieving and maintaining remission are vital in the management of the pregnant patient with IBD.

CONSIDERATIONS AFTER BIRTH: BREAST-FEEDING AND VACCINATION

Breast-feeding is considered safe. Minuscule amounts of infliximab or adalimumab are transferred in breast milk but are unlikely to result in systemic immune suppression in the infant.⁷

Live-attenuated vaccines should be avoided for the first 6 months in infants exposed to anti-TNF agents in utero.^1,7,11 All other vaccines, including hepatitis B virus vaccine, should be given according to standard schedules.⁶

OUR RECOMMENDATIONS

The goal of managing IBD in women of reproductive age is to minimize the risk of adverse outcomes for both mother and baby. We recommend a team approach, working closely with a gastroenterologist and a high-risk-pregnancy obstetrician, if available.

Patients should continue anti-TNF therapy during pregnancy because evidence supports its safety. If a woman wants to stop therapy and is at low risk of relapse, we recommend giving the last dose at 22 to 24 weeks of gestation, then promptly resuming therapy postpartum.

Live-attenuated vaccines (eg, influenza, rotavirus) should be avoided for the first 6 months in babies born to mothers on anti-TNF therapy.

No. Continuous monitoring for changes in heart rhythm with cardiac telemetry is recommended for all patients admitted to an intensive care unit (ICU). But routine telemetry monitoring for patients in non-ICU beds is not recommended, as it leads to unnecessary testing and treatment, increasing the cost of care and hospital length of stay.

RISK STRATIFICATION AND INDICATIONS

Telemetry is generally recommended for patients admitted with any type of heart disease, including:

• Acute myocardial infarction with ST-segment elevation or Q waves on 12-lead electrocardiography (ECG)
• Acute ischemia suggested by ST-segment depression or T-wave inversion on ECG
• Systolic blood pressure less than 100 mm Hg
• Acute decompensated heart failure with bilateral rales above the lung bases
• Chest pain that is worse than or the same as that in prior angina or myocardial infarction.^1,2

Indications for telemetry are less clear in patients with no history of heart disease. The American Heart Association (AHA)³ has classified admitted patients based on the presence or absence of heart disease³:

• Class I (high risk of arrhythmia): acute coronary syndrome, new arrhythmia (eg, atrial fibrillation or flutter), severe electrolyte imbalance; telemetry is warranted
• Class II (moderate risk): acute decompensated heart failure with stable hemodynamic status, a surgical or medical diagnosis with underlying paced rhythms (ie, with a pacemaker), and chronic arrhythmia (atrial fibrillation or flutter); in these cases, telemetry monitoring may be considered
• Class III (low risk): no history of cardiac disease or arrhythmias, admitted for medical or surgical reasons; in these cases, telemetry is generally not indicated³

Telemetry should also be considered in patients admitted with syncope or stroke, critical illness, or palpitations.

Syncope and stroke

Despite the wide use of telemetry for patients admitted with syncope, current evidence does not support this practice. However, the AHA recommends routine telemetry for patients admitted with idiopathic syncope when there is a high level of suspicion for underlying cardiac arrhythmias as a cause of syncope (risk class II-b).³ In 30% of patients admitted with stroke or transient ischemic attack, the cause is cardioembolic. Therefore, telemetry is indicated to rule out an underlying cardiac cause.⁴

Critical illness

Patients hospitalized with major trauma, acute respiratory failure, sepsis, shock, or acute pulmonary embolism or for major noncardiac surgery (especially elderly patients with coronary artery disease or at high risk of coronary events) require cardiac telemetry (risk class I-b). Patients admitted with kidney failure, significant electrolyte abnormalities, drug or substance toxicity (especially with known arrhythmogenic drugs) also require cardiac telemetry at the time of admission (risk class I-b).

Recurrent palpitations, arrhythmia

Most patients with palpitations can be evaluated in an outpatient setting.⁵ However, patients hospitalized for recurrent palpitations or for suspected underlying cardiac disease require telemetric monitoring (risk class II-b).³ Patients with high-degree atrioventricular block admitted after percutaneous temporary pacemaker implantation should be monitored, as should patients with a permanent pacemaker for 12 to 24 hours after implantation (risk class I-c). Also, patients hospitalized after implantable cardioverter-defibrillator (ICD) shock need to be monitored.^3,6

Patients with a paced rhythm who do not meet the above criteria do not require routine telemetric monitoring (risk class III-c).⁷

 

 

TELEMETRY IS OVERUSED

Off-site telemetry monitoring can identify significant arrhythmias during hospitalization. It also saves time on nursing staff to focus on bedside patient care. However, its convenience can lower the threshold for ordering it. This can lead to overuse with a major impact on healthcare costs.

Routine use of cardiac telemetry is associated with increased hospitalization costs with little benefit.⁸ The use of off-site services for continuous monitoring can activate many alarms throughout the day, triggering unnecessary workups and leading to densensitization to alarms (“alarm fatigue”).⁹

Despite the precise indications outlined in the AHA updated practice standards for inpatient electrocardiographic monitoring,¹⁰ telemetry use is expanding to non-ICU units without evidence of benefit,⁸ and this overuse can result in harmful clinical outcomes and a financial burden. Telemetry monitoring of low-risk patients can cause delays in emergency department and ICU admissions and transfers^8,11 of patients who may be sicker and need intensive care.

In a prospective observational study,¹² only 11 (6%) of 182 patients admitted to a general medical floor met AHA class I criteria for telemetry; very few patients developed a significant telemetry event such as atrial fibrillation or flutter that necessitated a change in management. Most overprescribers of telemetry monitoring reason that it will catch arrhythmias early.¹² In fact, in a study of patients in a cardiac unit, telemetry detected just 50% of in-house cardiac arrest cases, with a potential survival benefit of only 0.02%.¹³

Another study showed that only 0.01% of all telemetry alarms represented a real emergency. Only 37.2% of emergency alarms were classified as clinically important, and only 48.3% of these led to a change in management within 1 hour.¹⁴

Moreover, in a report of trauma patients with abnormal results on ECG at the time of admission, telemetry had negligible clinical benefit.¹⁵ And in a study of 414 patients, only 4% of those admitted with chest pain and normal initial ECG had cardiac interventions.¹⁶

Another study⁸ showed that hospital intervention to restrict the use of telemetry guided by AHA recommendations resulted in a 43% reduction in telemetry orders, a 47% reduction in telemetry duration, and a 70% reduction in the mean daily number of patients monitored, with no changes in hospital census or rates of code blue, death, or rapid response team activation.⁸

The financial cost can be seen in the backup of patients in the emergency department. A study showed that 91% of patients being admitted for chest pain were delayed by more than 3 hours while waiting for monitored beds. This translated into an annual cost of $168,300 to the hospital.¹⁷ Adherence to guidelines for appropriate use of telemetry can significantly decrease costs. Applying a simple algorithm for telemetry use was shown⁸ to decrease daily non-ICU cardiac telemetry costs from $18,971 to $5,772.

CURRENT GUIDELINES ARE LIMITED

The current American College of Cardiology and AHA guidelines are based mostly on expert opinion rather than randomized clinical trials, while most telemetry trials have been performed on patients with a cardiac or possible cardiac diagnosis.³ Current guidelines need to be updated, and more studies are needed to specify the optimal duration of cardiac monitoring in indicated cases. Many noncardiac conditions raise a legitimate concern of dysrhythmia, an indication for cardiac monitoring, but precise recommendations for telemetry for such conditions are lacking.

RECOMMENDATIONS


Raising awareness of the clinical and financial burdens associated with unwise telemetry utilization is critical. We suggest use of a pop-up notification in the electronic medical record to remind the provider of the existing telemetry order and to specify the duration of telemetry monitoring when placing the initial order. The goal is to identify patients in true need of a telemetry bed, to decrease unnecessary testing, and to reduce hospitalization costs.

No. Continuous monitoring for changes in heart rhythm with cardiac telemetry is recommended for all patients admitted to an intensive care unit (ICU). But routine telemetry monitoring for patients in non-ICU beds is not recommended, as it leads to unnecessary testing and treatment, increasing the cost of care and hospital length of stay.

RISK STRATIFICATION AND INDICATIONS

Telemetry is generally recommended for patients admitted with any type of heart disease, including:

• Acute myocardial infarction with ST-segment elevation or Q waves on 12-lead electrocardiography (ECG)
• Acute ischemia suggested by ST-segment depression or T-wave inversion on ECG
• Systolic blood pressure less than 100 mm Hg
• Acute decompensated heart failure with bilateral rales above the lung bases
• Chest pain that is worse than or the same as that in prior angina or myocardial infarction.^1,2

Indications for telemetry are less clear in patients with no history of heart disease. The American Heart Association (AHA)³ has classified admitted patients based on the presence or absence of heart disease³:

• Class I (high risk of arrhythmia): acute coronary syndrome, new arrhythmia (eg, atrial fibrillation or flutter), severe electrolyte imbalance; telemetry is warranted
• Class II (moderate risk): acute decompensated heart failure with stable hemodynamic status, a surgical or medical diagnosis with underlying paced rhythms (ie, with a pacemaker), and chronic arrhythmia (atrial fibrillation or flutter); in these cases, telemetry monitoring may be considered
• Class III (low risk): no history of cardiac disease or arrhythmias, admitted for medical or surgical reasons; in these cases, telemetry is generally not indicated³

Telemetry should also be considered in patients admitted with syncope or stroke, critical illness, or palpitations.

Syncope and stroke

Despite the wide use of telemetry for patients admitted with syncope, current evidence does not support this practice. However, the AHA recommends routine telemetry for patients admitted with idiopathic syncope when there is a high level of suspicion for underlying cardiac arrhythmias as a cause of syncope (risk class II-b).³ In 30% of patients admitted with stroke or transient ischemic attack, the cause is cardioembolic. Therefore, telemetry is indicated to rule out an underlying cardiac cause.⁴

Critical illness

Patients hospitalized with major trauma, acute respiratory failure, sepsis, shock, or acute pulmonary embolism or for major noncardiac surgery (especially elderly patients with coronary artery disease or at high risk of coronary events) require cardiac telemetry (risk class I-b). Patients admitted with kidney failure, significant electrolyte abnormalities, drug or substance toxicity (especially with known arrhythmogenic drugs) also require cardiac telemetry at the time of admission (risk class I-b).

Recurrent palpitations, arrhythmia

Most patients with palpitations can be evaluated in an outpatient setting.⁵ However, patients hospitalized for recurrent palpitations or for suspected underlying cardiac disease require telemetric monitoring (risk class II-b).³ Patients with high-degree atrioventricular block admitted after percutaneous temporary pacemaker implantation should be monitored, as should patients with a permanent pacemaker for 12 to 24 hours after implantation (risk class I-c). Also, patients hospitalized after implantable cardioverter-defibrillator (ICD) shock need to be monitored.^3,6

Patients with a paced rhythm who do not meet the above criteria do not require routine telemetric monitoring (risk class III-c).⁷

 

 

TELEMETRY IS OVERUSED

Off-site telemetry monitoring can identify significant arrhythmias during hospitalization. It also saves time on nursing staff to focus on bedside patient care. However, its convenience can lower the threshold for ordering it. This can lead to overuse with a major impact on healthcare costs.

Routine use of cardiac telemetry is associated with increased hospitalization costs with little benefit.⁸ The use of off-site services for continuous monitoring can activate many alarms throughout the day, triggering unnecessary workups and leading to densensitization to alarms (“alarm fatigue”).⁹

Despite the precise indications outlined in the AHA updated practice standards for inpatient electrocardiographic monitoring,¹⁰ telemetry use is expanding to non-ICU units without evidence of benefit,⁸ and this overuse can result in harmful clinical outcomes and a financial burden. Telemetry monitoring of low-risk patients can cause delays in emergency department and ICU admissions and transfers^8,11 of patients who may be sicker and need intensive care.

In a prospective observational study,¹² only 11 (6%) of 182 patients admitted to a general medical floor met AHA class I criteria for telemetry; very few patients developed a significant telemetry event such as atrial fibrillation or flutter that necessitated a change in management. Most overprescribers of telemetry monitoring reason that it will catch arrhythmias early.¹² In fact, in a study of patients in a cardiac unit, telemetry detected just 50% of in-house cardiac arrest cases, with a potential survival benefit of only 0.02%.¹³

Another study showed that only 0.01% of all telemetry alarms represented a real emergency. Only 37.2% of emergency alarms were classified as clinically important, and only 48.3% of these led to a change in management within 1 hour.¹⁴

Moreover, in a report of trauma patients with abnormal results on ECG at the time of admission, telemetry had negligible clinical benefit.¹⁵ And in a study of 414 patients, only 4% of those admitted with chest pain and normal initial ECG had cardiac interventions.¹⁶

Another study⁸ showed that hospital intervention to restrict the use of telemetry guided by AHA recommendations resulted in a 43% reduction in telemetry orders, a 47% reduction in telemetry duration, and a 70% reduction in the mean daily number of patients monitored, with no changes in hospital census or rates of code blue, death, or rapid response team activation.⁸

The financial cost can be seen in the backup of patients in the emergency department. A study showed that 91% of patients being admitted for chest pain were delayed by more than 3 hours while waiting for monitored beds. This translated into an annual cost of $168,300 to the hospital.¹⁷ Adherence to guidelines for appropriate use of telemetry can significantly decrease costs. Applying a simple algorithm for telemetry use was shown⁸ to decrease daily non-ICU cardiac telemetry costs from $18,971 to $5,772.

CURRENT GUIDELINES ARE LIMITED

The current American College of Cardiology and AHA guidelines are based mostly on expert opinion rather than randomized clinical trials, while most telemetry trials have been performed on patients with a cardiac or possible cardiac diagnosis.³ Current guidelines need to be updated, and more studies are needed to specify the optimal duration of cardiac monitoring in indicated cases. Many noncardiac conditions raise a legitimate concern of dysrhythmia, an indication for cardiac monitoring, but precise recommendations for telemetry for such conditions are lacking.

RECOMMENDATIONS


Raising awareness of the clinical and financial burdens associated with unwise telemetry utilization is critical. We suggest use of a pop-up notification in the electronic medical record to remind the provider of the existing telemetry order and to specify the duration of telemetry monitoring when placing the initial order. The goal is to identify patients in true need of a telemetry bed, to decrease unnecessary testing, and to reduce hospitalization costs.

No. Continuous monitoring for changes in heart rhythm with cardiac telemetry is recommended for all patients admitted to an intensive care unit (ICU). But routine telemetry monitoring for patients in non-ICU beds is not recommended, as it leads to unnecessary testing and treatment, increasing the cost of care and hospital length of stay.

RISK STRATIFICATION AND INDICATIONS

Telemetry is generally recommended for patients admitted with any type of heart disease, including:

• Acute myocardial infarction with ST-segment elevation or Q waves on 12-lead electrocardiography (ECG)
• Acute ischemia suggested by ST-segment depression or T-wave inversion on ECG
• Systolic blood pressure less than 100 mm Hg
• Acute decompensated heart failure with bilateral rales above the lung bases
• Chest pain that is worse than or the same as that in prior angina or myocardial infarction.^1,2

Indications for telemetry are less clear in patients with no history of heart disease. The American Heart Association (AHA)³ has classified admitted patients based on the presence or absence of heart disease³:

• Class I (high risk of arrhythmia): acute coronary syndrome, new arrhythmia (eg, atrial fibrillation or flutter), severe electrolyte imbalance; telemetry is warranted
• Class II (moderate risk): acute decompensated heart failure with stable hemodynamic status, a surgical or medical diagnosis with underlying paced rhythms (ie, with a pacemaker), and chronic arrhythmia (atrial fibrillation or flutter); in these cases, telemetry monitoring may be considered
• Class III (low risk): no history of cardiac disease or arrhythmias, admitted for medical or surgical reasons; in these cases, telemetry is generally not indicated³

Telemetry should also be considered in patients admitted with syncope or stroke, critical illness, or palpitations.

Syncope and stroke

Despite the wide use of telemetry for patients admitted with syncope, current evidence does not support this practice. However, the AHA recommends routine telemetry for patients admitted with idiopathic syncope when there is a high level of suspicion for underlying cardiac arrhythmias as a cause of syncope (risk class II-b).³ In 30% of patients admitted with stroke or transient ischemic attack, the cause is cardioembolic. Therefore, telemetry is indicated to rule out an underlying cardiac cause.⁴

Critical illness

Patients hospitalized with major trauma, acute respiratory failure, sepsis, shock, or acute pulmonary embolism or for major noncardiac surgery (especially elderly patients with coronary artery disease or at high risk of coronary events) require cardiac telemetry (risk class I-b). Patients admitted with kidney failure, significant electrolyte abnormalities, drug or substance toxicity (especially with known arrhythmogenic drugs) also require cardiac telemetry at the time of admission (risk class I-b).

Recurrent palpitations, arrhythmia

Most patients with palpitations can be evaluated in an outpatient setting.⁵ However, patients hospitalized for recurrent palpitations or for suspected underlying cardiac disease require telemetric monitoring (risk class II-b).³ Patients with high-degree atrioventricular block admitted after percutaneous temporary pacemaker implantation should be monitored, as should patients with a permanent pacemaker for 12 to 24 hours after implantation (risk class I-c). Also, patients hospitalized after implantable cardioverter-defibrillator (ICD) shock need to be monitored.^3,6

Patients with a paced rhythm who do not meet the above criteria do not require routine telemetric monitoring (risk class III-c).⁷

 

 

TELEMETRY IS OVERUSED

Off-site telemetry monitoring can identify significant arrhythmias during hospitalization. It also saves time on nursing staff to focus on bedside patient care. However, its convenience can lower the threshold for ordering it. This can lead to overuse with a major impact on healthcare costs.

Routine use of cardiac telemetry is associated with increased hospitalization costs with little benefit.⁸ The use of off-site services for continuous monitoring can activate many alarms throughout the day, triggering unnecessary workups and leading to densensitization to alarms (“alarm fatigue”).⁹

Despite the precise indications outlined in the AHA updated practice standards for inpatient electrocardiographic monitoring,¹⁰ telemetry use is expanding to non-ICU units without evidence of benefit,⁸ and this overuse can result in harmful clinical outcomes and a financial burden. Telemetry monitoring of low-risk patients can cause delays in emergency department and ICU admissions and transfers^8,11 of patients who may be sicker and need intensive care.

In a prospective observational study,¹² only 11 (6%) of 182 patients admitted to a general medical floor met AHA class I criteria for telemetry; very few patients developed a significant telemetry event such as atrial fibrillation or flutter that necessitated a change in management. Most overprescribers of telemetry monitoring reason that it will catch arrhythmias early.¹² In fact, in a study of patients in a cardiac unit, telemetry detected just 50% of in-house cardiac arrest cases, with a potential survival benefit of only 0.02%.¹³

Another study showed that only 0.01% of all telemetry alarms represented a real emergency. Only 37.2% of emergency alarms were classified as clinically important, and only 48.3% of these led to a change in management within 1 hour.¹⁴

Moreover, in a report of trauma patients with abnormal results on ECG at the time of admission, telemetry had negligible clinical benefit.¹⁵ And in a study of 414 patients, only 4% of those admitted with chest pain and normal initial ECG had cardiac interventions.¹⁶

Another study⁸ showed that hospital intervention to restrict the use of telemetry guided by AHA recommendations resulted in a 43% reduction in telemetry orders, a 47% reduction in telemetry duration, and a 70% reduction in the mean daily number of patients monitored, with no changes in hospital census or rates of code blue, death, or rapid response team activation.⁸

The financial cost can be seen in the backup of patients in the emergency department. A study showed that 91% of patients being admitted for chest pain were delayed by more than 3 hours while waiting for monitored beds. This translated into an annual cost of $168,300 to the hospital.¹⁷ Adherence to guidelines for appropriate use of telemetry can significantly decrease costs. Applying a simple algorithm for telemetry use was shown⁸ to decrease daily non-ICU cardiac telemetry costs from $18,971 to $5,772.

CURRENT GUIDELINES ARE LIMITED

The current American College of Cardiology and AHA guidelines are based mostly on expert opinion rather than randomized clinical trials, while most telemetry trials have been performed on patients with a cardiac or possible cardiac diagnosis.³ Current guidelines need to be updated, and more studies are needed to specify the optimal duration of cardiac monitoring in indicated cases. Many noncardiac conditions raise a legitimate concern of dysrhythmia, an indication for cardiac monitoring, but precise recommendations for telemetry for such conditions are lacking.

RECOMMENDATIONS


Raising awareness of the clinical and financial burdens associated with unwise telemetry utilization is critical. We suggest use of a pop-up notification in the electronic medical record to remind the provider of the existing telemetry order and to specify the duration of telemetry monitoring when placing the initial order. The goal is to identify patients in true need of a telemetry bed, to decrease unnecessary testing, and to reduce hospitalization costs.

Telemedicine has been used successfully to improve patient access to medical care while reducing healthcare costs. In 2016, an estimated 61% of US healthcare institutions and 40% to 50% of US hospitals used telemedicine.¹ From 2012 to 2013, the telemedicine market grew by 60%. However, its widespread use has been limited by low reimbursement rates and interstate licensing and practice issues.

In this commentary, we discuss the history of telemedicine, current uses and challenges, and areas of future growth.

DEFINITION AND HISTORY

The World Health Organization defines telemedicine as “the delivery of health care services, where distance is a critical factor, by all health care professionals using information and communication technologies for the exchange of valid information for diagnosis, treatment and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”²

Modern telemedicine began in the early 1900s in the Netherlands with the transmission of heart rhythms over the telephone,³ which was followed by transmissions to radio consultation centers in Europe in the 1920s. In the 1940s, radiographic images were transmitted by telephone between cities in Pennsylvania.⁴

Today, telemedicine is used in a variety of specialties including radiology, neurology, and pathology⁵ and by organizations in the United States ranging from the National Aeronautics and Space Administration and Kaiser Permanente to the US Department of Veterans Affairs (VA). The VA, in particular, is a leader in telemedicine. In 2012, it reduced mental health hospitalizations by over 40%, heart failure hospitalizations by 25%, and diabetes and chronic obstructive pulmonary disease hospitalizations by about 20% using telemedicine programs.⁶ In 2015, it provided about 2.1 million telemedicine consultations to 677,000 veterans.⁷

TYPES OF TELEMEDICINE PROGRAMS

There are 2 types of telemedicine programs.

Synchronous programs take place in real time and are a live 2-way interaction between the patient and healthcare provider. This includes virtual appointments that are conducted using the patient’s smartphone, tablet, or computer with a camera. When using a smartphone or tablet, patients must first download an app that connects them with a provider.

Asynchronous programs, also known as “store and forward” applications, are not live and involve the transfer of images, videos, and other clinical information that a healthcare provider views and responds to at a later time. In this case, patients may wear medical devices to monitor and track health information (eg, blood pressure) in a personal health application that they forward to their healthcare provider.

IMPROVING PATIENT ACCESS TO CARE WHILE REDUCING COSTS

Telemedicine allows patients living in both rural and urban areas to access healthcare when they need it. Currently, about 59 million Americans reside in health professional-shortage areas, which are rural and urban areas with shortages of primary care providers.¹ These patients often experience long delays when attempting to schedule a healthcare visit⁷ and may experience issues with continuity of care if they are unable to see the same care provider at every visit.

It also provides access to care to patients without reliable transportation or those who may be too sick to travel long distances. For some patients, such as those with cystic fibrosis who do not want to come to the hospital for fear of contracting multiple antibiotic-resistant bacteria, a virtual office visit may be safer.

At the same time, telemedicine helps reduce healthcare costs. For example, it:

• Optimizes staff distribution and healthcare resources within a healthcare facility and across an entire system
• Enables primary care providers to conduct appointments without additional office staff at any time, thereby extending office hours and availability
• Reduces the financial impact of patient no-shows
• Improves patient engagement and outcomes
• Reduces unnecessary office and emergency room visits and hospital admissions.

The last point is especially important for senior living and skilled nursing centers whose residents are known to have high rates of hospital admissions.^8,9 In these facilities, 24-hour medical assistance may not be available, and telemedicine can help troubleshoot common problems.

LOW REIMBURSEMENT RATES CURTAIL USE

Limited reimbursement has curtailed the widespread use of telemedicine. Although rules for reimbursement are evolving, telemedicine still represents a small amount of total healthcare expenditures. In 2015, Medicare spent approximately $14.4 million on services delivered via telemedicine—less than 0.01% of total spending on healthcare services.¹

Currently, 31 states and the District of Columbia have telemedicine parity laws that mandate private commercial insurers to pay for telemedicine services.¹⁰ Unfortunately, there is a lack of uniformity in the specifics of these laws, resulting in variations in reimbursement rates. Furthermore, a large number of larger insurers such as Medicare and Medicaid and many self-insured plans do not fall under these mandates.

Another factor that affects reimbursement for telemedicine services is the setting of the medical encounter. Medicare reimburses providers for telemedicine services only when they are conducted at specific sites such as physician’s offices, hospitals, rural health centers, and skilled nursing facilities. Additionally, Medicare only reimburses for services in areas with a shortage of healthcare professionals and in non-metropolitan areas, which excludes many urban patients.¹¹

In contrast, more commercial reimbursement is occurring for online urgent care, and options for commercial reimbursement of online behavioral services are being explored.

 

 

INTERSTATE LICENSURE ISSUES

Current licensure laws also limit the ability of many healthcare providers to offer telemedicine services. Federal law requires providers to be fully licensed to practice medicine in the state where the patient is physically located. In cases of health systems that have locations in more than one state, providers may need to apply for and pay to maintain multiple licenses (current interstate licensing laws vary across states).

Interstate licensure is one way to solve this problem. Thus far, a number of states have joined the Interstate Medical Licensure Compact that intends to allow physicians to obtain expedited licenses to practice in multiple states.¹²

The federal TELE-MED Act was introduced in 2015 but not passed. It proposed to “allow a Medicare provider to provide telemedicine services to a Medicare beneficiary who is in a different state from the one in which the provider is licensed or authorized to provide healthcare services.”

CAN TELEMEDICINE FOSTER A GOOD PROVIDER-PATIENT RELATIONSHIP?

In-person encounters provide healthcare providers with the opportunity to build a therapeutic relationship with their patients. Face-to-face encounters also increase patient satisfaction scores and outcomes. As such, critics fear that patient relationships may suffer with the use of telemedicine. However, using video technology for new patient encounters may help overcome this challenge. During a video encounter, the provider can see the patient’s facial expressions and take cues from nonverbal behaviors.

At times, the element of distance may enhance the encounter. For example, in behavioral health, patients often feel more comfortable in their home environment than in a sterile office environment.

Telemedicine patients often have positive experiences, given the speed of access, precision, time savings, and the ability to stay in contact with healthcare providers from the comfort of their homes. Ultimately, these virtual visits may help improve compliance with follow-up consultations since the barriers of distance and transportation are circumvented.

WHO CAN CONDUCT TELEMEDICINE VISITS?

Although a patient’s healthcare team is likely to consist of members who are not physicians, including nurse practitioners, physician assistants, social workers, and psychologists, not everyone can, by law, conduct telemedicine visits. Currently, the rules and regulations addressing ancillary team members’ participation in telemedicine vary from state to state.

TELEMEDICINE VISITS AT CLEVELAND CLINIC

Our health system has several telemedicine programs, including our eHospital program. Launched in 2014, this program provides patients at 4 hospitals with input from staff intensivists and experienced critical care nurses during the night (7 pm to 7 am) via remote monitoring. These remote caregivers have full access to patient charts and, when signalled, can activate an in-room camera to initiate 2-way audio communication with patients, their families, and bedside caregivers.

In addition, new patient consults are being offered via telemedicine for several services including dermatology, where pictures of skin lesions are reviewed and triaged, and management recommendations are provided accordingly.

In 2016, Cleveland Clinic launched its Remote Hypertension Improvement Program—an enterprise-wide initiative to minimize hypertension-associated mortality and morbidity with the assistance of telehealth services. The program was first piloted in a group of 80 high-risk hypertensive patients who were monitored and followed through a Bluetooth-enabled remote monitoring tool, which exported blood pressure readings to a central dashboard. A multidisciplinary team of doctors, nurses, and pharmacists used this dashboard to adjust medication when needed and provide virtual lifestyle coaching. Over a 24-week period, the patients’ systolic blood pressure decreased by an average of 7.5 mm Hg and diastolic blood pressure by 3.1 mm Hg (unpublished data).

Beginning this year, blood pressure readings will be directly exported from the remote monitoring tool into the patient’s electronic medical record, providing the healthcare team with the information needed to make informed decisions to remotely manage patients with hypertension.

Remote monitoring of patients with hypertension is also being used at other institutions such as the VA. In 2016, almost 19,000 veterans were using the remote monitoring system, and this number is expected to increase with the enhanced adaptation of telemedicine services.¹³

FUTURE DIRECTIONS

About 50% of all adults in the United States have at least 1 chronic disease. In all, chronic disease accounts for roughly 75% of the total healthcare expenditure and 70% of all deaths.^7,14 Recent data suggest that virtual chronic disease management represents an untapped market for telemedicine, given its relative underutilization compared to other services such as telebehavorial health and specialty telemedicine. These patients require frequent visits to the doctor, and targeting this patient population with telemedicine may decrease the number of emergency room visits and hospital admissions.

Another growing area in the field of telemedicine is the “hospital at home” model in which patients who meet the criteria for hospitalization but are otherwise stable are treated at home for diseases such as chronic obstructive pulmonary disease, pneumonia, and heart failure. Studies have shown that the hospital-at-home model, when used appropriately, is not only more cost-effective than hospitalization but results in a shorter treatment duration and lower rates of delirium.^15–17

Finally, in the acute setting, we have seen wide success with telemedicine programs in stroke care, radiology, intensive care, and psychiatry, and several studies have shown mortality rates comparable to those with the traditional model.^18,19 These encounters often require specialized skills and are the focus of multiple ongoing studies.           

Acknowledgment: The authors would like to acknowledge and thank Matthew Faiman, MD, for providing information regarding the Remote Hypertension Program.

Telemedicine has been used successfully to improve patient access to medical care while reducing healthcare costs. In 2016, an estimated 61% of US healthcare institutions and 40% to 50% of US hospitals used telemedicine.¹ From 2012 to 2013, the telemedicine market grew by 60%. However, its widespread use has been limited by low reimbursement rates and interstate licensing and practice issues.

In this commentary, we discuss the history of telemedicine, current uses and challenges, and areas of future growth.

DEFINITION AND HISTORY

The World Health Organization defines telemedicine as “the delivery of health care services, where distance is a critical factor, by all health care professionals using information and communication technologies for the exchange of valid information for diagnosis, treatment and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”²

Modern telemedicine began in the early 1900s in the Netherlands with the transmission of heart rhythms over the telephone,³ which was followed by transmissions to radio consultation centers in Europe in the 1920s. In the 1940s, radiographic images were transmitted by telephone between cities in Pennsylvania.⁴

Today, telemedicine is used in a variety of specialties including radiology, neurology, and pathology⁵ and by organizations in the United States ranging from the National Aeronautics and Space Administration and Kaiser Permanente to the US Department of Veterans Affairs (VA). The VA, in particular, is a leader in telemedicine. In 2012, it reduced mental health hospitalizations by over 40%, heart failure hospitalizations by 25%, and diabetes and chronic obstructive pulmonary disease hospitalizations by about 20% using telemedicine programs.⁶ In 2015, it provided about 2.1 million telemedicine consultations to 677,000 veterans.⁷

TYPES OF TELEMEDICINE PROGRAMS

There are 2 types of telemedicine programs.

Synchronous programs take place in real time and are a live 2-way interaction between the patient and healthcare provider. This includes virtual appointments that are conducted using the patient’s smartphone, tablet, or computer with a camera. When using a smartphone or tablet, patients must first download an app that connects them with a provider.

Asynchronous programs, also known as “store and forward” applications, are not live and involve the transfer of images, videos, and other clinical information that a healthcare provider views and responds to at a later time. In this case, patients may wear medical devices to monitor and track health information (eg, blood pressure) in a personal health application that they forward to their healthcare provider.

IMPROVING PATIENT ACCESS TO CARE WHILE REDUCING COSTS

Telemedicine allows patients living in both rural and urban areas to access healthcare when they need it. Currently, about 59 million Americans reside in health professional-shortage areas, which are rural and urban areas with shortages of primary care providers.¹ These patients often experience long delays when attempting to schedule a healthcare visit⁷ and may experience issues with continuity of care if they are unable to see the same care provider at every visit.

It also provides access to care to patients without reliable transportation or those who may be too sick to travel long distances. For some patients, such as those with cystic fibrosis who do not want to come to the hospital for fear of contracting multiple antibiotic-resistant bacteria, a virtual office visit may be safer.

At the same time, telemedicine helps reduce healthcare costs. For example, it:

• Optimizes staff distribution and healthcare resources within a healthcare facility and across an entire system
• Enables primary care providers to conduct appointments without additional office staff at any time, thereby extending office hours and availability
• Reduces the financial impact of patient no-shows
• Improves patient engagement and outcomes
• Reduces unnecessary office and emergency room visits and hospital admissions.

The last point is especially important for senior living and skilled nursing centers whose residents are known to have high rates of hospital admissions.^8,9 In these facilities, 24-hour medical assistance may not be available, and telemedicine can help troubleshoot common problems.

LOW REIMBURSEMENT RATES CURTAIL USE

Limited reimbursement has curtailed the widespread use of telemedicine. Although rules for reimbursement are evolving, telemedicine still represents a small amount of total healthcare expenditures. In 2015, Medicare spent approximately $14.4 million on services delivered via telemedicine—less than 0.01% of total spending on healthcare services.¹

Currently, 31 states and the District of Columbia have telemedicine parity laws that mandate private commercial insurers to pay for telemedicine services.¹⁰ Unfortunately, there is a lack of uniformity in the specifics of these laws, resulting in variations in reimbursement rates. Furthermore, a large number of larger insurers such as Medicare and Medicaid and many self-insured plans do not fall under these mandates.

Another factor that affects reimbursement for telemedicine services is the setting of the medical encounter. Medicare reimburses providers for telemedicine services only when they are conducted at specific sites such as physician’s offices, hospitals, rural health centers, and skilled nursing facilities. Additionally, Medicare only reimburses for services in areas with a shortage of healthcare professionals and in non-metropolitan areas, which excludes many urban patients.¹¹

In contrast, more commercial reimbursement is occurring for online urgent care, and options for commercial reimbursement of online behavioral services are being explored.

 

 

INTERSTATE LICENSURE ISSUES

Current licensure laws also limit the ability of many healthcare providers to offer telemedicine services. Federal law requires providers to be fully licensed to practice medicine in the state where the patient is physically located. In cases of health systems that have locations in more than one state, providers may need to apply for and pay to maintain multiple licenses (current interstate licensing laws vary across states).

Interstate licensure is one way to solve this problem. Thus far, a number of states have joined the Interstate Medical Licensure Compact that intends to allow physicians to obtain expedited licenses to practice in multiple states.¹²

The federal TELE-MED Act was introduced in 2015 but not passed. It proposed to “allow a Medicare provider to provide telemedicine services to a Medicare beneficiary who is in a different state from the one in which the provider is licensed or authorized to provide healthcare services.”

CAN TELEMEDICINE FOSTER A GOOD PROVIDER-PATIENT RELATIONSHIP?

In-person encounters provide healthcare providers with the opportunity to build a therapeutic relationship with their patients. Face-to-face encounters also increase patient satisfaction scores and outcomes. As such, critics fear that patient relationships may suffer with the use of telemedicine. However, using video technology for new patient encounters may help overcome this challenge. During a video encounter, the provider can see the patient’s facial expressions and take cues from nonverbal behaviors.

At times, the element of distance may enhance the encounter. For example, in behavioral health, patients often feel more comfortable in their home environment than in a sterile office environment.

Telemedicine patients often have positive experiences, given the speed of access, precision, time savings, and the ability to stay in contact with healthcare providers from the comfort of their homes. Ultimately, these virtual visits may help improve compliance with follow-up consultations since the barriers of distance and transportation are circumvented.

WHO CAN CONDUCT TELEMEDICINE VISITS?

Although a patient’s healthcare team is likely to consist of members who are not physicians, including nurse practitioners, physician assistants, social workers, and psychologists, not everyone can, by law, conduct telemedicine visits. Currently, the rules and regulations addressing ancillary team members’ participation in telemedicine vary from state to state.

TELEMEDICINE VISITS AT CLEVELAND CLINIC

Our health system has several telemedicine programs, including our eHospital program. Launched in 2014, this program provides patients at 4 hospitals with input from staff intensivists and experienced critical care nurses during the night (7 pm to 7 am) via remote monitoring. These remote caregivers have full access to patient charts and, when signalled, can activate an in-room camera to initiate 2-way audio communication with patients, their families, and bedside caregivers.

In addition, new patient consults are being offered via telemedicine for several services including dermatology, where pictures of skin lesions are reviewed and triaged, and management recommendations are provided accordingly.

In 2016, Cleveland Clinic launched its Remote Hypertension Improvement Program—an enterprise-wide initiative to minimize hypertension-associated mortality and morbidity with the assistance of telehealth services. The program was first piloted in a group of 80 high-risk hypertensive patients who were monitored and followed through a Bluetooth-enabled remote monitoring tool, which exported blood pressure readings to a central dashboard. A multidisciplinary team of doctors, nurses, and pharmacists used this dashboard to adjust medication when needed and provide virtual lifestyle coaching. Over a 24-week period, the patients’ systolic blood pressure decreased by an average of 7.5 mm Hg and diastolic blood pressure by 3.1 mm Hg (unpublished data).

Beginning this year, blood pressure readings will be directly exported from the remote monitoring tool into the patient’s electronic medical record, providing the healthcare team with the information needed to make informed decisions to remotely manage patients with hypertension.

Remote monitoring of patients with hypertension is also being used at other institutions such as the VA. In 2016, almost 19,000 veterans were using the remote monitoring system, and this number is expected to increase with the enhanced adaptation of telemedicine services.¹³

FUTURE DIRECTIONS

About 50% of all adults in the United States have at least 1 chronic disease. In all, chronic disease accounts for roughly 75% of the total healthcare expenditure and 70% of all deaths.^7,14 Recent data suggest that virtual chronic disease management represents an untapped market for telemedicine, given its relative underutilization compared to other services such as telebehavorial health and specialty telemedicine. These patients require frequent visits to the doctor, and targeting this patient population with telemedicine may decrease the number of emergency room visits and hospital admissions.

Another growing area in the field of telemedicine is the “hospital at home” model in which patients who meet the criteria for hospitalization but are otherwise stable are treated at home for diseases such as chronic obstructive pulmonary disease, pneumonia, and heart failure. Studies have shown that the hospital-at-home model, when used appropriately, is not only more cost-effective than hospitalization but results in a shorter treatment duration and lower rates of delirium.^15–17

Finally, in the acute setting, we have seen wide success with telemedicine programs in stroke care, radiology, intensive care, and psychiatry, and several studies have shown mortality rates comparable to those with the traditional model.^18,19 These encounters often require specialized skills and are the focus of multiple ongoing studies.           

Acknowledgment: The authors would like to acknowledge and thank Matthew Faiman, MD, for providing information regarding the Remote Hypertension Program.

Telemedicine has been used successfully to improve patient access to medical care while reducing healthcare costs. In 2016, an estimated 61% of US healthcare institutions and 40% to 50% of US hospitals used telemedicine.¹ From 2012 to 2013, the telemedicine market grew by 60%. However, its widespread use has been limited by low reimbursement rates and interstate licensing and practice issues.

In this commentary, we discuss the history of telemedicine, current uses and challenges, and areas of future growth.

DEFINITION AND HISTORY

The World Health Organization defines telemedicine as “the delivery of health care services, where distance is a critical factor, by all health care professionals using information and communication technologies for the exchange of valid information for diagnosis, treatment and prevention of disease and injuries, research and evaluation, and for the continuing education of healthcare providers, all in the interests of advancing the health of individuals and their communities.”²

Modern telemedicine began in the early 1900s in the Netherlands with the transmission of heart rhythms over the telephone,³ which was followed by transmissions to radio consultation centers in Europe in the 1920s. In the 1940s, radiographic images were transmitted by telephone between cities in Pennsylvania.⁴

Today, telemedicine is used in a variety of specialties including radiology, neurology, and pathology⁵ and by organizations in the United States ranging from the National Aeronautics and Space Administration and Kaiser Permanente to the US Department of Veterans Affairs (VA). The VA, in particular, is a leader in telemedicine. In 2012, it reduced mental health hospitalizations by over 40%, heart failure hospitalizations by 25%, and diabetes and chronic obstructive pulmonary disease hospitalizations by about 20% using telemedicine programs.⁶ In 2015, it provided about 2.1 million telemedicine consultations to 677,000 veterans.⁷

TYPES OF TELEMEDICINE PROGRAMS

There are 2 types of telemedicine programs.

Synchronous programs take place in real time and are a live 2-way interaction between the patient and healthcare provider. This includes virtual appointments that are conducted using the patient’s smartphone, tablet, or computer with a camera. When using a smartphone or tablet, patients must first download an app that connects them with a provider.

Asynchronous programs, also known as “store and forward” applications, are not live and involve the transfer of images, videos, and other clinical information that a healthcare provider views and responds to at a later time. In this case, patients may wear medical devices to monitor and track health information (eg, blood pressure) in a personal health application that they forward to their healthcare provider.

IMPROVING PATIENT ACCESS TO CARE WHILE REDUCING COSTS

Telemedicine allows patients living in both rural and urban areas to access healthcare when they need it. Currently, about 59 million Americans reside in health professional-shortage areas, which are rural and urban areas with shortages of primary care providers.¹ These patients often experience long delays when attempting to schedule a healthcare visit⁷ and may experience issues with continuity of care if they are unable to see the same care provider at every visit.

It also provides access to care to patients without reliable transportation or those who may be too sick to travel long distances. For some patients, such as those with cystic fibrosis who do not want to come to the hospital for fear of contracting multiple antibiotic-resistant bacteria, a virtual office visit may be safer.

At the same time, telemedicine helps reduce healthcare costs. For example, it:

• Optimizes staff distribution and healthcare resources within a healthcare facility and across an entire system
• Enables primary care providers to conduct appointments without additional office staff at any time, thereby extending office hours and availability
• Reduces the financial impact of patient no-shows
• Improves patient engagement and outcomes
• Reduces unnecessary office and emergency room visits and hospital admissions.

The last point is especially important for senior living and skilled nursing centers whose residents are known to have high rates of hospital admissions.^8,9 In these facilities, 24-hour medical assistance may not be available, and telemedicine can help troubleshoot common problems.

LOW REIMBURSEMENT RATES CURTAIL USE

Limited reimbursement has curtailed the widespread use of telemedicine. Although rules for reimbursement are evolving, telemedicine still represents a small amount of total healthcare expenditures. In 2015, Medicare spent approximately $14.4 million on services delivered via telemedicine—less than 0.01% of total spending on healthcare services.¹

Currently, 31 states and the District of Columbia have telemedicine parity laws that mandate private commercial insurers to pay for telemedicine services.¹⁰ Unfortunately, there is a lack of uniformity in the specifics of these laws, resulting in variations in reimbursement rates. Furthermore, a large number of larger insurers such as Medicare and Medicaid and many self-insured plans do not fall under these mandates.

Another factor that affects reimbursement for telemedicine services is the setting of the medical encounter. Medicare reimburses providers for telemedicine services only when they are conducted at specific sites such as physician’s offices, hospitals, rural health centers, and skilled nursing facilities. Additionally, Medicare only reimburses for services in areas with a shortage of healthcare professionals and in non-metropolitan areas, which excludes many urban patients.¹¹

In contrast, more commercial reimbursement is occurring for online urgent care, and options for commercial reimbursement of online behavioral services are being explored.

 

 

INTERSTATE LICENSURE ISSUES

Current licensure laws also limit the ability of many healthcare providers to offer telemedicine services. Federal law requires providers to be fully licensed to practice medicine in the state where the patient is physically located. In cases of health systems that have locations in more than one state, providers may need to apply for and pay to maintain multiple licenses (current interstate licensing laws vary across states).

Interstate licensure is one way to solve this problem. Thus far, a number of states have joined the Interstate Medical Licensure Compact that intends to allow physicians to obtain expedited licenses to practice in multiple states.¹²

The federal TELE-MED Act was introduced in 2015 but not passed. It proposed to “allow a Medicare provider to provide telemedicine services to a Medicare beneficiary who is in a different state from the one in which the provider is licensed or authorized to provide healthcare services.”

CAN TELEMEDICINE FOSTER A GOOD PROVIDER-PATIENT RELATIONSHIP?

In-person encounters provide healthcare providers with the opportunity to build a therapeutic relationship with their patients. Face-to-face encounters also increase patient satisfaction scores and outcomes. As such, critics fear that patient relationships may suffer with the use of telemedicine. However, using video technology for new patient encounters may help overcome this challenge. During a video encounter, the provider can see the patient’s facial expressions and take cues from nonverbal behaviors.

At times, the element of distance may enhance the encounter. For example, in behavioral health, patients often feel more comfortable in their home environment than in a sterile office environment.

Telemedicine patients often have positive experiences, given the speed of access, precision, time savings, and the ability to stay in contact with healthcare providers from the comfort of their homes. Ultimately, these virtual visits may help improve compliance with follow-up consultations since the barriers of distance and transportation are circumvented.

WHO CAN CONDUCT TELEMEDICINE VISITS?

Although a patient’s healthcare team is likely to consist of members who are not physicians, including nurse practitioners, physician assistants, social workers, and psychologists, not everyone can, by law, conduct telemedicine visits. Currently, the rules and regulations addressing ancillary team members’ participation in telemedicine vary from state to state.

TELEMEDICINE VISITS AT CLEVELAND CLINIC

Our health system has several telemedicine programs, including our eHospital program. Launched in 2014, this program provides patients at 4 hospitals with input from staff intensivists and experienced critical care nurses during the night (7 pm to 7 am) via remote monitoring. These remote caregivers have full access to patient charts and, when signalled, can activate an in-room camera to initiate 2-way audio communication with patients, their families, and bedside caregivers.

In addition, new patient consults are being offered via telemedicine for several services including dermatology, where pictures of skin lesions are reviewed and triaged, and management recommendations are provided accordingly.

In 2016, Cleveland Clinic launched its Remote Hypertension Improvement Program—an enterprise-wide initiative to minimize hypertension-associated mortality and morbidity with the assistance of telehealth services. The program was first piloted in a group of 80 high-risk hypertensive patients who were monitored and followed through a Bluetooth-enabled remote monitoring tool, which exported blood pressure readings to a central dashboard. A multidisciplinary team of doctors, nurses, and pharmacists used this dashboard to adjust medication when needed and provide virtual lifestyle coaching. Over a 24-week period, the patients’ systolic blood pressure decreased by an average of 7.5 mm Hg and diastolic blood pressure by 3.1 mm Hg (unpublished data).

Beginning this year, blood pressure readings will be directly exported from the remote monitoring tool into the patient’s electronic medical record, providing the healthcare team with the information needed to make informed decisions to remotely manage patients with hypertension.

Remote monitoring of patients with hypertension is also being used at other institutions such as the VA. In 2016, almost 19,000 veterans were using the remote monitoring system, and this number is expected to increase with the enhanced adaptation of telemedicine services.¹³

FUTURE DIRECTIONS

About 50% of all adults in the United States have at least 1 chronic disease. In all, chronic disease accounts for roughly 75% of the total healthcare expenditure and 70% of all deaths.^7,14 Recent data suggest that virtual chronic disease management represents an untapped market for telemedicine, given its relative underutilization compared to other services such as telebehavorial health and specialty telemedicine. These patients require frequent visits to the doctor, and targeting this patient population with telemedicine may decrease the number of emergency room visits and hospital admissions.

Another growing area in the field of telemedicine is the “hospital at home” model in which patients who meet the criteria for hospitalization but are otherwise stable are treated at home for diseases such as chronic obstructive pulmonary disease, pneumonia, and heart failure. Studies have shown that the hospital-at-home model, when used appropriately, is not only more cost-effective than hospitalization but results in a shorter treatment duration and lower rates of delirium.^15–17

Finally, in the acute setting, we have seen wide success with telemedicine programs in stroke care, radiology, intensive care, and psychiatry, and several studies have shown mortality rates comparable to those with the traditional model.^18,19 These encounters often require specialized skills and are the focus of multiple ongoing studies.           

Acknowledgment: The authors would like to acknowledge and thank Matthew Faiman, MD, for providing information regarding the Remote Hypertension Program.