We recommend the following combination treatment:

Normal saline (0.9% NaCl) 1 to 2 L by intravenous (IV) infusion over 2 to 4 hours.  This can be repeated every 6 to 12 hours.

Ketorolac 30-mg IV bolus, which can be repeated every 6 hours. However, patients with coronary artery disease, uncontrolled hypertension, acute renal failure, or cerebrovascular disease should instead receive aceta­minophen 1,000 mg, naproxen sodium 550 mg, or aspirin 325 mg by mouth.

Prochlorperazine or metoclopramide 10-mg IV infusion. This can be repeated every 6 hours. However, to reduce the extrapyramidal adverse effects of these drugs, patients should first receive diphenhydramine 25- to 50-mg IV bolus, which can be repeated every 6 to 8 hours.

Sodium valproate 500 to 1,000 mg by IV infusion over 20 minutes. This can be repeated after 8 hours.

Dexamethasone 4-mg IV bolus every 6 hours, or 10-mg IV bolus once in 24 hours.

Magnesium sulfate 500 to 1,000 mg by IV infusion over 1 hour. This can be repeated every 6 to 12 hours.

If the migraine has not improved after 3 cycles of this regimen, a neurologic consultation should be considered. Other options include dihydroergotamine and occipital nerve blocks¹ performed at the bedside.

GENERAL PRINCIPLES

Managing acute intractable migraine can be frustrating for both the practitioner and the patient. Some general principles are helpful.

Use a combination of drugs. Aborting a severe migraine attack often requires a combination of medications that work synergistically.

Use IV and intramuscular formulations rather than oral formulations: they are more rapidly absorbed, provide faster pain relief, and can be given when the nausea that often accompanies migraine precludes oral treatments.

Rule out secondary causes. The mnemonic SNOOP—systemic signs, neurologic signs, onset, older age, progression of existing headache disorder—is useful for assessing underlying causes.² Any patient presenting with intractable migraine should also have a thorough neuro­logic examination.

Screening electrocardiography may be helpful, as the pretreatment QTc interval may direct the choice of intravenous treatment. If the patient has a prolonged QTc or is taking another drug that could prolong the QTc, certain medications, specifically dopamine receptor antagonists and diphenhydra­mine, should be avoided.

Ask the patient what has worked previously. A particular agent may have been effective in aborting the migraine; thus, a single dose of it could abort the headache, expediting discharge.

Establish if a triptan or ergot derivative has been used during the 24 hours before presentation, as repeated dosing within this interval is not recommended.³

Establish the baseline headache severity. Complete headache relief is difficult to achieve in a patient with chronic daily headache, and establishing a more realistic goal (eg, 50% relief) from the outset is useful.

Antiemetics

Dopamine receptor antagonists are assumed to merely treat nausea in patients with migraine; however, they act independently to abort migraine and thus should be considered, irrespective of the presence of nausea.

The two most commonly used agents are prochlorperazine and metoclopramide. The American Academy of Neurology guidelines recommend prochlorperazine as first-line therapy for acute migraine. Metoclopramide is rated slightly lower and is considered to have moderate benefit.⁴ The Canadian Headache Society cites a high level of evidence supporting prochlorperazine and a moderate level of evidence supporting metoclopramide.⁵ The American Headache Society assessment of parenteral pharmacotherapies gives prochlorperazine and metoclopramide a level B recommendation of “should offer” (a recommendation only additionally assigned to subcutaneous sumatriptan).³ Hence, either agent can be used.

To reduce the risk of posttreatment akathisia, diphenhydramine or benztropine may be given before starting a dopamine receptor antagonist. Diphenhydramine may be independently effective in migraine treatment,^6,7 but  data on this are limited.

Ketorolac, ibuprofen

Ketorolac and ibuprofen are the only available nonsteroidal antiinflammatory drugs (NSAIDs) for IV administration. The Canadian Headache Society guidelines strongly recommend ketorolac for the treatment of migraine in emergency settings.⁵ Doses range from 30 mg to 60 mg.¹ Ibuprofen 400 to 800 mg by IV infusion is an acceptable alternative. These medications should be avoided in patients with renal failure or severe coronary artery disease.

Oral naproxen sodium is a possible alternative in patients with cardiovascular disease, as it has been shown to carry a lower cardiovascular risk than other NSAIDs.⁸

The same concerns in patients with renal dysfunction apply to any NSAID, as the enzyme cyclooxygenase plays a constitutive role in glomerular function.

Antiepileptic drugs

The antiepileptic drugs sodium valproate and levetiracetam are available in IV formulations that have demonstrated efficacy in the treatment of status migrainosus¹ (ie, migraine lasting more than 72 hours). Valproate has the strongest track record, is well tolerated, and is effective in rapidly aborting migraine.⁹

Volume repletion

Although its use is anecdotal and to date no trial has measured its efficacy, IV volume repletion is often used in acute migraine, as most headache experts surmise it to be highly effective, especially in patients with prolonged nausea or vomiting.¹

Magnesium

IV magnesium is effective, particularly for migraine with aura.¹⁰ Hypotension is a common side effect, and pretreatment or concurrent treatment with IV fluids is advised. Doses from 500 mg to 1,000 mg have demonstrated efficacy.¹⁰

Corticosteroids

Corticosteroids can be used in the treatment of status migrainosus. Most studies have shown benefit in preventing recurrences rather than merely aborting migraine.¹¹ A systematic review suggested that recurrent headaches are milder with corticosteroid treatment; 19 of 25 studies indicated favorable benefit, and 6 of 19 studies indicated noninferior outcomes.¹²

Both IV methylprednisolone and IV dexamethasone may be considered.¹² Dexamethasone appears to be particularly effective in preventing headache recurrence when combined with other IV therapies.¹³ It can be given as a single dose of 10 mg, or as repeated doses of 4 mg up to 16 mg/day.¹ Patients with active psychosis or uncontrolled diabetes should be closely monitored for these conditions, which corticosteroids can worsen.

Serotoninergic agents

Serotonin agonists including subcutaneous sumatriptan and IV dihydroergotamine are highly effective, with proven statistical and clinical benefit.⁴ They should be considered in patients with no known history of coronary artery disease or other vaso-occlusive vascular disorder.¹

Ideally, IV dihydroergotamine should be administered after consultation with a neurologist or headache specialist, given the pretreatment and cotreatment requirements often necessary to suppress its side effects. Careful titration is important to prevent transient headache exacerbations during infusion, as well as abdominal cramping, nausea, and diarrhea.

Avoid opioids

Opioids should be avoided. Evidence supporting their use in acute migraine is extremely limited,³ and the risks of migraine becoming chronic and of addiction are high.¹⁴ Safer, more effective alternatives have been detailed above.

A detailed algorithm for the management of patients with acute migraine has been published¹⁴ and is aimed at decreasing acute treatment with opioids and barbiturates.

We recommend the following combination treatment:

Normal saline (0.9% NaCl) 1 to 2 L by intravenous (IV) infusion over 2 to 4 hours.  This can be repeated every 6 to 12 hours.

Ketorolac 30-mg IV bolus, which can be repeated every 6 hours. However, patients with coronary artery disease, uncontrolled hypertension, acute renal failure, or cerebrovascular disease should instead receive aceta­minophen 1,000 mg, naproxen sodium 550 mg, or aspirin 325 mg by mouth.

Prochlorperazine or metoclopramide 10-mg IV infusion. This can be repeated every 6 hours. However, to reduce the extrapyramidal adverse effects of these drugs, patients should first receive diphenhydramine 25- to 50-mg IV bolus, which can be repeated every 6 to 8 hours.

Sodium valproate 500 to 1,000 mg by IV infusion over 20 minutes. This can be repeated after 8 hours.

Dexamethasone 4-mg IV bolus every 6 hours, or 10-mg IV bolus once in 24 hours.

Magnesium sulfate 500 to 1,000 mg by IV infusion over 1 hour. This can be repeated every 6 to 12 hours.

If the migraine has not improved after 3 cycles of this regimen, a neurologic consultation should be considered. Other options include dihydroergotamine and occipital nerve blocks¹ performed at the bedside.

GENERAL PRINCIPLES

Managing acute intractable migraine can be frustrating for both the practitioner and the patient. Some general principles are helpful.

Use a combination of drugs. Aborting a severe migraine attack often requires a combination of medications that work synergistically.

Use IV and intramuscular formulations rather than oral formulations: they are more rapidly absorbed, provide faster pain relief, and can be given when the nausea that often accompanies migraine precludes oral treatments.

Rule out secondary causes. The mnemonic SNOOP—systemic signs, neurologic signs, onset, older age, progression of existing headache disorder—is useful for assessing underlying causes.² Any patient presenting with intractable migraine should also have a thorough neuro­logic examination.

Screening electrocardiography may be helpful, as the pretreatment QTc interval may direct the choice of intravenous treatment. If the patient has a prolonged QTc or is taking another drug that could prolong the QTc, certain medications, specifically dopamine receptor antagonists and diphenhydra­mine, should be avoided.

Ask the patient what has worked previously. A particular agent may have been effective in aborting the migraine; thus, a single dose of it could abort the headache, expediting discharge.

Establish if a triptan or ergot derivative has been used during the 24 hours before presentation, as repeated dosing within this interval is not recommended.³

Establish the baseline headache severity. Complete headache relief is difficult to achieve in a patient with chronic daily headache, and establishing a more realistic goal (eg, 50% relief) from the outset is useful.

Antiemetics

Dopamine receptor antagonists are assumed to merely treat nausea in patients with migraine; however, they act independently to abort migraine and thus should be considered, irrespective of the presence of nausea.

The two most commonly used agents are prochlorperazine and metoclopramide. The American Academy of Neurology guidelines recommend prochlorperazine as first-line therapy for acute migraine. Metoclopramide is rated slightly lower and is considered to have moderate benefit.⁴ The Canadian Headache Society cites a high level of evidence supporting prochlorperazine and a moderate level of evidence supporting metoclopramide.⁵ The American Headache Society assessment of parenteral pharmacotherapies gives prochlorperazine and metoclopramide a level B recommendation of “should offer” (a recommendation only additionally assigned to subcutaneous sumatriptan).³ Hence, either agent can be used.

To reduce the risk of posttreatment akathisia, diphenhydramine or benztropine may be given before starting a dopamine receptor antagonist. Diphenhydramine may be independently effective in migraine treatment,^6,7 but  data on this are limited.

Ketorolac, ibuprofen

Ketorolac and ibuprofen are the only available nonsteroidal antiinflammatory drugs (NSAIDs) for IV administration. The Canadian Headache Society guidelines strongly recommend ketorolac for the treatment of migraine in emergency settings.⁵ Doses range from 30 mg to 60 mg.¹ Ibuprofen 400 to 800 mg by IV infusion is an acceptable alternative. These medications should be avoided in patients with renal failure or severe coronary artery disease.

Oral naproxen sodium is a possible alternative in patients with cardiovascular disease, as it has been shown to carry a lower cardiovascular risk than other NSAIDs.⁸

The same concerns in patients with renal dysfunction apply to any NSAID, as the enzyme cyclooxygenase plays a constitutive role in glomerular function.

Antiepileptic drugs

The antiepileptic drugs sodium valproate and levetiracetam are available in IV formulations that have demonstrated efficacy in the treatment of status migrainosus¹ (ie, migraine lasting more than 72 hours). Valproate has the strongest track record, is well tolerated, and is effective in rapidly aborting migraine.⁹

Volume repletion

Although its use is anecdotal and to date no trial has measured its efficacy, IV volume repletion is often used in acute migraine, as most headache experts surmise it to be highly effective, especially in patients with prolonged nausea or vomiting.¹

Magnesium

IV magnesium is effective, particularly for migraine with aura.¹⁰ Hypotension is a common side effect, and pretreatment or concurrent treatment with IV fluids is advised. Doses from 500 mg to 1,000 mg have demonstrated efficacy.¹⁰

Corticosteroids

Corticosteroids can be used in the treatment of status migrainosus. Most studies have shown benefit in preventing recurrences rather than merely aborting migraine.¹¹ A systematic review suggested that recurrent headaches are milder with corticosteroid treatment; 19 of 25 studies indicated favorable benefit, and 6 of 19 studies indicated noninferior outcomes.¹²

Both IV methylprednisolone and IV dexamethasone may be considered.¹² Dexamethasone appears to be particularly effective in preventing headache recurrence when combined with other IV therapies.¹³ It can be given as a single dose of 10 mg, or as repeated doses of 4 mg up to 16 mg/day.¹ Patients with active psychosis or uncontrolled diabetes should be closely monitored for these conditions, which corticosteroids can worsen.

Serotoninergic agents

Serotonin agonists including subcutaneous sumatriptan and IV dihydroergotamine are highly effective, with proven statistical and clinical benefit.⁴ They should be considered in patients with no known history of coronary artery disease or other vaso-occlusive vascular disorder.¹

Ideally, IV dihydroergotamine should be administered after consultation with a neurologist or headache specialist, given the pretreatment and cotreatment requirements often necessary to suppress its side effects. Careful titration is important to prevent transient headache exacerbations during infusion, as well as abdominal cramping, nausea, and diarrhea.

Avoid opioids

Opioids should be avoided. Evidence supporting their use in acute migraine is extremely limited,³ and the risks of migraine becoming chronic and of addiction are high.¹⁴ Safer, more effective alternatives have been detailed above.

A detailed algorithm for the management of patients with acute migraine has been published¹⁴ and is aimed at decreasing acute treatment with opioids and barbiturates.

We recommend the following combination treatment:

Normal saline (0.9% NaCl) 1 to 2 L by intravenous (IV) infusion over 2 to 4 hours.  This can be repeated every 6 to 12 hours.

Ketorolac 30-mg IV bolus, which can be repeated every 6 hours. However, patients with coronary artery disease, uncontrolled hypertension, acute renal failure, or cerebrovascular disease should instead receive aceta­minophen 1,000 mg, naproxen sodium 550 mg, or aspirin 325 mg by mouth.

Prochlorperazine or metoclopramide 10-mg IV infusion. This can be repeated every 6 hours. However, to reduce the extrapyramidal adverse effects of these drugs, patients should first receive diphenhydramine 25- to 50-mg IV bolus, which can be repeated every 6 to 8 hours.

Sodium valproate 500 to 1,000 mg by IV infusion over 20 minutes. This can be repeated after 8 hours.

Dexamethasone 4-mg IV bolus every 6 hours, or 10-mg IV bolus once in 24 hours.

Magnesium sulfate 500 to 1,000 mg by IV infusion over 1 hour. This can be repeated every 6 to 12 hours.

If the migraine has not improved after 3 cycles of this regimen, a neurologic consultation should be considered. Other options include dihydroergotamine and occipital nerve blocks¹ performed at the bedside.

GENERAL PRINCIPLES

Managing acute intractable migraine can be frustrating for both the practitioner and the patient. Some general principles are helpful.

Use a combination of drugs. Aborting a severe migraine attack often requires a combination of medications that work synergistically.

Use IV and intramuscular formulations rather than oral formulations: they are more rapidly absorbed, provide faster pain relief, and can be given when the nausea that often accompanies migraine precludes oral treatments.

Rule out secondary causes. The mnemonic SNOOP—systemic signs, neurologic signs, onset, older age, progression of existing headache disorder—is useful for assessing underlying causes.² Any patient presenting with intractable migraine should also have a thorough neuro­logic examination.

Screening electrocardiography may be helpful, as the pretreatment QTc interval may direct the choice of intravenous treatment. If the patient has a prolonged QTc or is taking another drug that could prolong the QTc, certain medications, specifically dopamine receptor antagonists and diphenhydra­mine, should be avoided.

Ask the patient what has worked previously. A particular agent may have been effective in aborting the migraine; thus, a single dose of it could abort the headache, expediting discharge.

Establish if a triptan or ergot derivative has been used during the 24 hours before presentation, as repeated dosing within this interval is not recommended.³

Establish the baseline headache severity. Complete headache relief is difficult to achieve in a patient with chronic daily headache, and establishing a more realistic goal (eg, 50% relief) from the outset is useful.

Antiemetics

Dopamine receptor antagonists are assumed to merely treat nausea in patients with migraine; however, they act independently to abort migraine and thus should be considered, irrespective of the presence of nausea.

The two most commonly used agents are prochlorperazine and metoclopramide. The American Academy of Neurology guidelines recommend prochlorperazine as first-line therapy for acute migraine. Metoclopramide is rated slightly lower and is considered to have moderate benefit.⁴ The Canadian Headache Society cites a high level of evidence supporting prochlorperazine and a moderate level of evidence supporting metoclopramide.⁵ The American Headache Society assessment of parenteral pharmacotherapies gives prochlorperazine and metoclopramide a level B recommendation of “should offer” (a recommendation only additionally assigned to subcutaneous sumatriptan).³ Hence, either agent can be used.

To reduce the risk of posttreatment akathisia, diphenhydramine or benztropine may be given before starting a dopamine receptor antagonist. Diphenhydramine may be independently effective in migraine treatment,^6,7 but  data on this are limited.

Ketorolac, ibuprofen

Ketorolac and ibuprofen are the only available nonsteroidal antiinflammatory drugs (NSAIDs) for IV administration. The Canadian Headache Society guidelines strongly recommend ketorolac for the treatment of migraine in emergency settings.⁵ Doses range from 30 mg to 60 mg.¹ Ibuprofen 400 to 800 mg by IV infusion is an acceptable alternative. These medications should be avoided in patients with renal failure or severe coronary artery disease.

Oral naproxen sodium is a possible alternative in patients with cardiovascular disease, as it has been shown to carry a lower cardiovascular risk than other NSAIDs.⁸

The same concerns in patients with renal dysfunction apply to any NSAID, as the enzyme cyclooxygenase plays a constitutive role in glomerular function.

Antiepileptic drugs

The antiepileptic drugs sodium valproate and levetiracetam are available in IV formulations that have demonstrated efficacy in the treatment of status migrainosus¹ (ie, migraine lasting more than 72 hours). Valproate has the strongest track record, is well tolerated, and is effective in rapidly aborting migraine.⁹

Volume repletion

Although its use is anecdotal and to date no trial has measured its efficacy, IV volume repletion is often used in acute migraine, as most headache experts surmise it to be highly effective, especially in patients with prolonged nausea or vomiting.¹

Magnesium

IV magnesium is effective, particularly for migraine with aura.¹⁰ Hypotension is a common side effect, and pretreatment or concurrent treatment with IV fluids is advised. Doses from 500 mg to 1,000 mg have demonstrated efficacy.¹⁰

Corticosteroids

Corticosteroids can be used in the treatment of status migrainosus. Most studies have shown benefit in preventing recurrences rather than merely aborting migraine.¹¹ A systematic review suggested that recurrent headaches are milder with corticosteroid treatment; 19 of 25 studies indicated favorable benefit, and 6 of 19 studies indicated noninferior outcomes.¹²

Both IV methylprednisolone and IV dexamethasone may be considered.¹² Dexamethasone appears to be particularly effective in preventing headache recurrence when combined with other IV therapies.¹³ It can be given as a single dose of 10 mg, or as repeated doses of 4 mg up to 16 mg/day.¹ Patients with active psychosis or uncontrolled diabetes should be closely monitored for these conditions, which corticosteroids can worsen.

Serotoninergic agents

Serotonin agonists including subcutaneous sumatriptan and IV dihydroergotamine are highly effective, with proven statistical and clinical benefit.⁴ They should be considered in patients with no known history of coronary artery disease or other vaso-occlusive vascular disorder.¹

Ideally, IV dihydroergotamine should be administered after consultation with a neurologist or headache specialist, given the pretreatment and cotreatment requirements often necessary to suppress its side effects. Careful titration is important to prevent transient headache exacerbations during infusion, as well as abdominal cramping, nausea, and diarrhea.

Avoid opioids

Opioids should be avoided. Evidence supporting their use in acute migraine is extremely limited,³ and the risks of migraine becoming chronic and of addiction are high.¹⁴ Safer, more effective alternatives have been detailed above.

A detailed algorithm for the management of patients with acute migraine has been published¹⁴ and is aimed at decreasing acute treatment with opioids and barbiturates.

Staphylococcus aureus is the most common infective agent in native and prosthetic valve endocarditis, and 13% to 22% of patients with S aureus bacteremia have infective endocarditis.¹

See related editorial

Transthoracic echocardiography (TTE) is a good starting point in the workup of suspected infective endocarditis, but transesophageal echocardiography (TEE) plays a key role in diagnosis and is indicated in patients with a high pretest probability of infective endocarditis, as in the following scenarios:

• Clinical picture consistent with infective endocarditis
• Presence of previously placed port or other indwelling vascular device
• Presence of a prosthetic valve or other prosthetic material 
• Presence of a pacemaker
• History of valve disease
• Injection drug use
• Positive blood cultures after 72 hours despite appropriate antibiotic treatment
• Abnormal TTE result requiring better visualization of valvular anatomy and function and confirmation of local complications
• Absence of another reasonable explanation for S aureus bacteremia.

Forgoing TEE is reasonable in patients with normal results on TTE, no predisposing risk factors, a reasonable alternative explanation for S aureus bacteremia, and a low pretest probability of infective endocarditis.¹ TEE may also be unnecessary if there is another disease focus requiring extended treatment (eg, vertebral infection) and there are no findings suggesting complicated infective endocarditis, eg, persistent bacteremia, symptoms of heart failure, and conduction abnormality.¹

TEE also may be unnecessary in patients at low risk who have identifiable foci of bacteremia due to soft-tissue infection or a newly placed vascular catheter and whose bacteremia clears within 72 hours of the start of antibiotic therapy. These patients may be followed clinically for the development of new findings such as metastatic foci of infection (eg, septic pulmonary emboli, renal infarction, splenic abscess or infarction), the new onset of heart failure or cardiac conduction abnormality, or recurrence of previously cleared S aureus bacteremia. If these should develop, then a more invasive study such as TEE may be warranted.

INFECTIVE ENDOCARDITIS: EPIDEMIOLOGY AND MICROBIOLOGY

The US incidence rate of infective endocarditis has steadily increased, with an estimated 457,052 hospitalizations from 2000 to 2011. During that period, from 2000 to 2007, there was a marked increase in valve replacement surgeries.² This trend is likely explained by an increase in the at-risk population—eg, elderly patients, patients with opiate dependence or diabetes, and patients on hemodialysis.

Although S aureus is the predominant pathogen in infective endocarditis,^2–5S aureus bacteremia is often observed in patients with skin or soft-tissue infection, prosthetic device infection, vascular graft or catheter infection, and bone and joint infections. S aureus bacteremia necessitates a search for the source of infection.

S aureus is a major pathogen in bloodstream infections, and up to 14% of patients with S aureus bacteremia have infective endocarditis as the primary source of infection.³ The pathogenesis of S aureus infective endocarditis is thought to be mediated by cell-wall factors that promote adhesion to the extracellular matrix of intravascular structures.³

A new localizing symptom such as back pain, joint pain, or swelling in a patient with S aureus bacteremia should trigger an investigation for metastatic infection.

Infectious disease consultation in patients with S aureus bacteremia is associated with improved outcomes and, thus, should be pursued.³

A cardiac surgery consult is recommended early on in cases of infective endocarditis caused by vancomycin-resistant enterococci, Pseudomonas aeruginosa, and fungi, as well as in patients with complications such as valvular insufficiency, perivalvular abscess, conduction abnormalities, persistent bacteremia, and metastatic foci of infection.⁶

Risk factors for infective endocarditis include injection drug abuse, valvular heart disease, congenital heart disease (unrepaired, repaired with residual defects, or fully repaired within the past 6 months), previous infective endocarditis, prosthetic heart valve, and cardiac transplant.^2–4,6 Other risk factors are poor dentition, hemodialysis, ventriculoatrial shunts, intravascular devices including vascular grafts, and pacemakers.^2,3 Many risk factors for infective endocarditis and S aureus bacteremia overlap.³

DIAGNOSTIC PRINCIPLES

The clinical presentation of infective endocarditis can vary from a nonspecific infectious syndrome, to overt organ failure (heart failure, kidney failure), to an acute vascular catastrophe (arterial ischemia, cerebrovascular accidents, myocardial infarction). Patients may present with indolent symptoms such as fever, fatigue, and weight loss,⁶ or they may present at an advanced stage, with fulminant acute heart failure due to valvular insufficiency or with arrhythmias due to a perivalvular abscess infiltrating the conduction system. Extracardiac clinical manifestations may be related to direct infective metastatic foci such as septic emboli or to immunologic phenomena such as glomerulonephritis or Osler nodes.

ECHOCARDIOGRAPHY’S ROLE IN DIAGNOSIS

TTE plays an important role in diagnosis and risk stratification of infective endocarditis.⁶ TTE is usually done first because of its low cost, wide availability, and safety; it has a sensitivity of 70% and a specificity over 95%.⁸ While a normal result on TTE does not completely rule out infective endocarditis, completely normal valvular morphology and function on TTE make the diagnosis less likely.^8,9

If suspicion remains high despite a normal study, repeating TTE at a later time may result in a higher diagnostic yield because of growth of the suspected vegetation. Otherwise, TEE should be considered.

TEE provides a higher spatial resolution and diagnostic yield than TTE, especially for detecting complex pathology such as pseudoaneurysm, valve perforation, or valvular abscess. TEE has a sensitivity and specificity of approximately 95% for infective endocarditis.⁸ It should be performed early in patients with preexisting valve disease, prosthetic cardiac material (eg, valves), or a pacemaker or implantable cardioverter-defibrillator.^6,7

Detecting valve vegetation provides answers about the cause of S aureus bacteremia with its complications (eg, septic emboli, mycotic aneurysm) and informs decisions about the duration of antibiotic therapy and the need for surgery.^3,6

As with any diagnostic test, it is important to compare the results of any recent study with those of previous studies whenever possible to differentiate new from old findings.

WHEN TO FORGO TEE IN S AUREUS BACTEREMIA

Because TEE is invasive and requires the patient to swallow an endoscopic probe,¹⁰ it is important to screen patients for esophageal disease, cervical spine conditions, and baseline respiratory insufficiency. Complications are rare but include esophageal perforation, esophageal bleeding, pharyngeal hematoma, and reactions to anesthesia.¹⁰

As with any diagnostic test, the clinician first needs to consider the patient’s pretest probability of the disease, the diagnostic accuracy, the associated risks and costs, and the implications of the results.

While TEE provides better diagnostic images than TTE, a normal TEE study does not exclude the diagnosis of infective endocarditis: small lesions and complications such as paravalvular abscess of a prosthetic aortic valve may still be missed. In such patients, a repeat TEE examination or additional imaging study (eg, gated computed tomographic angiography) should be considered.⁶

Noninfective sterile echodensities, valvular tumors such as papillary fibroelastomas, Lambl excrescences, and suture lines of prosthetic valves are among the conditions and factors that can cause a false-positive result on TEE. 

Staphylococcus aureus is the most common infective agent in native and prosthetic valve endocarditis, and 13% to 22% of patients with S aureus bacteremia have infective endocarditis.¹

See related editorial

Transthoracic echocardiography (TTE) is a good starting point in the workup of suspected infective endocarditis, but transesophageal echocardiography (TEE) plays a key role in diagnosis and is indicated in patients with a high pretest probability of infective endocarditis, as in the following scenarios:

• Clinical picture consistent with infective endocarditis
• Presence of previously placed port or other indwelling vascular device
• Presence of a prosthetic valve or other prosthetic material 
• Presence of a pacemaker
• History of valve disease
• Injection drug use
• Positive blood cultures after 72 hours despite appropriate antibiotic treatment
• Abnormal TTE result requiring better visualization of valvular anatomy and function and confirmation of local complications
• Absence of another reasonable explanation for S aureus bacteremia.

Forgoing TEE is reasonable in patients with normal results on TTE, no predisposing risk factors, a reasonable alternative explanation for S aureus bacteremia, and a low pretest probability of infective endocarditis.¹ TEE may also be unnecessary if there is another disease focus requiring extended treatment (eg, vertebral infection) and there are no findings suggesting complicated infective endocarditis, eg, persistent bacteremia, symptoms of heart failure, and conduction abnormality.¹

TEE also may be unnecessary in patients at low risk who have identifiable foci of bacteremia due to soft-tissue infection or a newly placed vascular catheter and whose bacteremia clears within 72 hours of the start of antibiotic therapy. These patients may be followed clinically for the development of new findings such as metastatic foci of infection (eg, septic pulmonary emboli, renal infarction, splenic abscess or infarction), the new onset of heart failure or cardiac conduction abnormality, or recurrence of previously cleared S aureus bacteremia. If these should develop, then a more invasive study such as TEE may be warranted.

INFECTIVE ENDOCARDITIS: EPIDEMIOLOGY AND MICROBIOLOGY

The US incidence rate of infective endocarditis has steadily increased, with an estimated 457,052 hospitalizations from 2000 to 2011. During that period, from 2000 to 2007, there was a marked increase in valve replacement surgeries.² This trend is likely explained by an increase in the at-risk population—eg, elderly patients, patients with opiate dependence or diabetes, and patients on hemodialysis.

Although S aureus is the predominant pathogen in infective endocarditis,^2–5S aureus bacteremia is often observed in patients with skin or soft-tissue infection, prosthetic device infection, vascular graft or catheter infection, and bone and joint infections. S aureus bacteremia necessitates a search for the source of infection.

S aureus is a major pathogen in bloodstream infections, and up to 14% of patients with S aureus bacteremia have infective endocarditis as the primary source of infection.³ The pathogenesis of S aureus infective endocarditis is thought to be mediated by cell-wall factors that promote adhesion to the extracellular matrix of intravascular structures.³

A new localizing symptom such as back pain, joint pain, or swelling in a patient with S aureus bacteremia should trigger an investigation for metastatic infection.

Infectious disease consultation in patients with S aureus bacteremia is associated with improved outcomes and, thus, should be pursued.³

A cardiac surgery consult is recommended early on in cases of infective endocarditis caused by vancomycin-resistant enterococci, Pseudomonas aeruginosa, and fungi, as well as in patients with complications such as valvular insufficiency, perivalvular abscess, conduction abnormalities, persistent bacteremia, and metastatic foci of infection.⁶

Risk factors for infective endocarditis include injection drug abuse, valvular heart disease, congenital heart disease (unrepaired, repaired with residual defects, or fully repaired within the past 6 months), previous infective endocarditis, prosthetic heart valve, and cardiac transplant.^2–4,6 Other risk factors are poor dentition, hemodialysis, ventriculoatrial shunts, intravascular devices including vascular grafts, and pacemakers.^2,3 Many risk factors for infective endocarditis and S aureus bacteremia overlap.³

DIAGNOSTIC PRINCIPLES

The clinical presentation of infective endocarditis can vary from a nonspecific infectious syndrome, to overt organ failure (heart failure, kidney failure), to an acute vascular catastrophe (arterial ischemia, cerebrovascular accidents, myocardial infarction). Patients may present with indolent symptoms such as fever, fatigue, and weight loss,⁶ or they may present at an advanced stage, with fulminant acute heart failure due to valvular insufficiency or with arrhythmias due to a perivalvular abscess infiltrating the conduction system. Extracardiac clinical manifestations may be related to direct infective metastatic foci such as septic emboli or to immunologic phenomena such as glomerulonephritis or Osler nodes.

ECHOCARDIOGRAPHY’S ROLE IN DIAGNOSIS

TTE plays an important role in diagnosis and risk stratification of infective endocarditis.⁶ TTE is usually done first because of its low cost, wide availability, and safety; it has a sensitivity of 70% and a specificity over 95%.⁸ While a normal result on TTE does not completely rule out infective endocarditis, completely normal valvular morphology and function on TTE make the diagnosis less likely.^8,9

If suspicion remains high despite a normal study, repeating TTE at a later time may result in a higher diagnostic yield because of growth of the suspected vegetation. Otherwise, TEE should be considered.

TEE provides a higher spatial resolution and diagnostic yield than TTE, especially for detecting complex pathology such as pseudoaneurysm, valve perforation, or valvular abscess. TEE has a sensitivity and specificity of approximately 95% for infective endocarditis.⁸ It should be performed early in patients with preexisting valve disease, prosthetic cardiac material (eg, valves), or a pacemaker or implantable cardioverter-defibrillator.^6,7

Detecting valve vegetation provides answers about the cause of S aureus bacteremia with its complications (eg, septic emboli, mycotic aneurysm) and informs decisions about the duration of antibiotic therapy and the need for surgery.^3,6

As with any diagnostic test, it is important to compare the results of any recent study with those of previous studies whenever possible to differentiate new from old findings.

WHEN TO FORGO TEE IN S AUREUS BACTEREMIA

Because TEE is invasive and requires the patient to swallow an endoscopic probe,¹⁰ it is important to screen patients for esophageal disease, cervical spine conditions, and baseline respiratory insufficiency. Complications are rare but include esophageal perforation, esophageal bleeding, pharyngeal hematoma, and reactions to anesthesia.¹⁰

As with any diagnostic test, the clinician first needs to consider the patient’s pretest probability of the disease, the diagnostic accuracy, the associated risks and costs, and the implications of the results.

While TEE provides better diagnostic images than TTE, a normal TEE study does not exclude the diagnosis of infective endocarditis: small lesions and complications such as paravalvular abscess of a prosthetic aortic valve may still be missed. In such patients, a repeat TEE examination or additional imaging study (eg, gated computed tomographic angiography) should be considered.⁶

Noninfective sterile echodensities, valvular tumors such as papillary fibroelastomas, Lambl excrescences, and suture lines of prosthetic valves are among the conditions and factors that can cause a false-positive result on TEE. 

Staphylococcus aureus is the most common infective agent in native and prosthetic valve endocarditis, and 13% to 22% of patients with S aureus bacteremia have infective endocarditis.¹

See related editorial

Transthoracic echocardiography (TTE) is a good starting point in the workup of suspected infective endocarditis, but transesophageal echocardiography (TEE) plays a key role in diagnosis and is indicated in patients with a high pretest probability of infective endocarditis, as in the following scenarios:

• Clinical picture consistent with infective endocarditis
• Presence of previously placed port or other indwelling vascular device
• Presence of a prosthetic valve or other prosthetic material 
• Presence of a pacemaker
• History of valve disease
• Injection drug use
• Positive blood cultures after 72 hours despite appropriate antibiotic treatment
• Abnormal TTE result requiring better visualization of valvular anatomy and function and confirmation of local complications
• Absence of another reasonable explanation for S aureus bacteremia.

Forgoing TEE is reasonable in patients with normal results on TTE, no predisposing risk factors, a reasonable alternative explanation for S aureus bacteremia, and a low pretest probability of infective endocarditis.¹ TEE may also be unnecessary if there is another disease focus requiring extended treatment (eg, vertebral infection) and there are no findings suggesting complicated infective endocarditis, eg, persistent bacteremia, symptoms of heart failure, and conduction abnormality.¹

TEE also may be unnecessary in patients at low risk who have identifiable foci of bacteremia due to soft-tissue infection or a newly placed vascular catheter and whose bacteremia clears within 72 hours of the start of antibiotic therapy. These patients may be followed clinically for the development of new findings such as metastatic foci of infection (eg, septic pulmonary emboli, renal infarction, splenic abscess or infarction), the new onset of heart failure or cardiac conduction abnormality, or recurrence of previously cleared S aureus bacteremia. If these should develop, then a more invasive study such as TEE may be warranted.

INFECTIVE ENDOCARDITIS: EPIDEMIOLOGY AND MICROBIOLOGY

The US incidence rate of infective endocarditis has steadily increased, with an estimated 457,052 hospitalizations from 2000 to 2011. During that period, from 2000 to 2007, there was a marked increase in valve replacement surgeries.² This trend is likely explained by an increase in the at-risk population—eg, elderly patients, patients with opiate dependence or diabetes, and patients on hemodialysis.

Although S aureus is the predominant pathogen in infective endocarditis,^2–5S aureus bacteremia is often observed in patients with skin or soft-tissue infection, prosthetic device infection, vascular graft or catheter infection, and bone and joint infections. S aureus bacteremia necessitates a search for the source of infection.

S aureus is a major pathogen in bloodstream infections, and up to 14% of patients with S aureus bacteremia have infective endocarditis as the primary source of infection.³ The pathogenesis of S aureus infective endocarditis is thought to be mediated by cell-wall factors that promote adhesion to the extracellular matrix of intravascular structures.³

A new localizing symptom such as back pain, joint pain, or swelling in a patient with S aureus bacteremia should trigger an investigation for metastatic infection.

Infectious disease consultation in patients with S aureus bacteremia is associated with improved outcomes and, thus, should be pursued.³

A cardiac surgery consult is recommended early on in cases of infective endocarditis caused by vancomycin-resistant enterococci, Pseudomonas aeruginosa, and fungi, as well as in patients with complications such as valvular insufficiency, perivalvular abscess, conduction abnormalities, persistent bacteremia, and metastatic foci of infection.⁶

Risk factors for infective endocarditis include injection drug abuse, valvular heart disease, congenital heart disease (unrepaired, repaired with residual defects, or fully repaired within the past 6 months), previous infective endocarditis, prosthetic heart valve, and cardiac transplant.^2–4,6 Other risk factors are poor dentition, hemodialysis, ventriculoatrial shunts, intravascular devices including vascular grafts, and pacemakers.^2,3 Many risk factors for infective endocarditis and S aureus bacteremia overlap.³

DIAGNOSTIC PRINCIPLES

The clinical presentation of infective endocarditis can vary from a nonspecific infectious syndrome, to overt organ failure (heart failure, kidney failure), to an acute vascular catastrophe (arterial ischemia, cerebrovascular accidents, myocardial infarction). Patients may present with indolent symptoms such as fever, fatigue, and weight loss,⁶ or they may present at an advanced stage, with fulminant acute heart failure due to valvular insufficiency or with arrhythmias due to a perivalvular abscess infiltrating the conduction system. Extracardiac clinical manifestations may be related to direct infective metastatic foci such as septic emboli or to immunologic phenomena such as glomerulonephritis or Osler nodes.

ECHOCARDIOGRAPHY’S ROLE IN DIAGNOSIS

TTE plays an important role in diagnosis and risk stratification of infective endocarditis.⁶ TTE is usually done first because of its low cost, wide availability, and safety; it has a sensitivity of 70% and a specificity over 95%.⁸ While a normal result on TTE does not completely rule out infective endocarditis, completely normal valvular morphology and function on TTE make the diagnosis less likely.^8,9

If suspicion remains high despite a normal study, repeating TTE at a later time may result in a higher diagnostic yield because of growth of the suspected vegetation. Otherwise, TEE should be considered.

TEE provides a higher spatial resolution and diagnostic yield than TTE, especially for detecting complex pathology such as pseudoaneurysm, valve perforation, or valvular abscess. TEE has a sensitivity and specificity of approximately 95% for infective endocarditis.⁸ It should be performed early in patients with preexisting valve disease, prosthetic cardiac material (eg, valves), or a pacemaker or implantable cardioverter-defibrillator.^6,7

Detecting valve vegetation provides answers about the cause of S aureus bacteremia with its complications (eg, septic emboli, mycotic aneurysm) and informs decisions about the duration of antibiotic therapy and the need for surgery.^3,6

As with any diagnostic test, it is important to compare the results of any recent study with those of previous studies whenever possible to differentiate new from old findings.

WHEN TO FORGO TEE IN S AUREUS BACTEREMIA

Because TEE is invasive and requires the patient to swallow an endoscopic probe,¹⁰ it is important to screen patients for esophageal disease, cervical spine conditions, and baseline respiratory insufficiency. Complications are rare but include esophageal perforation, esophageal bleeding, pharyngeal hematoma, and reactions to anesthesia.¹⁰

As with any diagnostic test, the clinician first needs to consider the patient’s pretest probability of the disease, the diagnostic accuracy, the associated risks and costs, and the implications of the results.

While TEE provides better diagnostic images than TTE, a normal TEE study does not exclude the diagnosis of infective endocarditis: small lesions and complications such as paravalvular abscess of a prosthetic aortic valve may still be missed. In such patients, a repeat TEE examination or additional imaging study (eg, gated computed tomographic angiography) should be considered.⁶

Noninfective sterile echodensities, valvular tumors such as papillary fibroelastomas, Lambl excrescences, and suture lines of prosthetic valves are among the conditions and factors that can cause a false-positive result on TEE. 

Morbidity and mortality rates in patients with Staphylococcus aureus bacteremia remain high even though diagnostic tests have improved and antibiotic therapy is effective. Diagnosis and management are made more complex by difficulties in finding the source of bacteremia and sites of metastatic infection.

See related article

S aureus bacteremia is a finding that demands further investigation, since up to 25% of people who have it may have endocarditis, a condition with even worse consequences.¹ The ability of S aureus to infect normal valves^2,3 adds to the challenge. In the mid-20th century, Wilson and Hamburger⁴ demonstrated that 64% of patients with S aureus bacteremia had evidence of valvular infection at autopsy. In a more recent case series of patients with S aureus endocarditis, the diagnosis was established at autopsy in 32%.⁵

Specific clinical findings in patients with complicated S aureus bacteremia—those who have a site of infection remote from or extended beyond the primary focus—may be useful in determining the need for additional diagnostic and therapeutic measures.

In a prospective cohort study, Fowler et al⁶ identified several factors that predicted complicated S aureus bacteremia (including but not limited to endocarditis):

• Prolonged bacteremia (> 48–72 hours after initiation of therapy)
• Community onset
• Fever persisting more than 72 hours
• Skin findings suggesting systemic infection.

THE ROLE OF ECHOCARDIOGRAPHY

Infective endocarditis may be difficult to detect in patients with S aureus bacteremia; experts recommend routine use of echocardiography in this process.^7,8 Transesophageal echocardiography (TEE) detects more cases of endocarditis than transthoracic echocardiography (TTE),^9,10 but access, cost, and risks lead to questions about its utility.

Guidance for the use of echocardiography in S aureus bacteremia^1,10–14 continues to evolve. Consensus seems to be emerging that the risk of endocarditis is lower in patients with S aureus bacteremia who:

• Do not have a prosthetic valve or other permanent intracardiac device
• Have sterile blood cultures within 96 hours after the initial set
• Are not hemodialysis-dependent
• Developed the bacteremia in a healthcare setting
• Have no secondary focus of infection
• Have no clinical signs of infective endocarditis.

Heriot et al¹⁴ point out that studies of risk-stratification approaches to echocardiography in patients with S aureus bacteremia are difficult to interpret, as there are questions regarding the validity of the studies and the balance of the risks and benefits.¹ The question of timing of TEE remains largely unexplored, both in initial screening and in follow-up of previously undiagnosed cases of S aureus endocarditis.

In this issue of the Journal, Mirrakhimov et al¹⁵ weigh in on use of a risk-stratification model to guide use of TEE in patients with S aureus bacteremia. Their comments about avoiding TEE in patients who have an alternative explanation for S aureus bacteremia and a low pretest probability for infectious endocarditis and in patients with a disease focus that requires extended treatment are derived from a survey of infectious disease physicians.¹⁶

ROLE OF INFECTIOUS DISEASE CONSULTATION

Infectious disease consultation reduces mortality rates and healthcare costs for a variety of infections, with endocarditis as a prime example.¹⁷ For S aureus bacteremia, a large and growing body of literature demonstrates the impact of infectious disease consultation, including improved adherence to guidelines and quality measures,^18–20 lower in-hospital mortality rates^18–21 and earlier hospital discharge.¹⁸ In the era of “curbside consults” and “e-consultation,” it is interesting to note the enduring value of bedside, in-person consultation in the management of S aureus bacteremia.²⁰

Many people with S aureus bacteremia should undergo TEE. Until the evidence becomes more robust, the decision to forgo TEE must be made with caution. The expertise of infectious disease physicians in the diagnosis and management of endocarditis can assist clinicians working with the often-complex patients who develop S aureus bacteremia. If the goal is to improve outcomes, infectious disease consultation may be at least as important as appropriate selection of patients for TEE.

Morbidity and mortality rates in patients with Staphylococcus aureus bacteremia remain high even though diagnostic tests have improved and antibiotic therapy is effective. Diagnosis and management are made more complex by difficulties in finding the source of bacteremia and sites of metastatic infection.

See related article

S aureus bacteremia is a finding that demands further investigation, since up to 25% of people who have it may have endocarditis, a condition with even worse consequences.¹ The ability of S aureus to infect normal valves^2,3 adds to the challenge. In the mid-20th century, Wilson and Hamburger⁴ demonstrated that 64% of patients with S aureus bacteremia had evidence of valvular infection at autopsy. In a more recent case series of patients with S aureus endocarditis, the diagnosis was established at autopsy in 32%.⁵

Specific clinical findings in patients with complicated S aureus bacteremia—those who have a site of infection remote from or extended beyond the primary focus—may be useful in determining the need for additional diagnostic and therapeutic measures.

In a prospective cohort study, Fowler et al⁶ identified several factors that predicted complicated S aureus bacteremia (including but not limited to endocarditis):

• Prolonged bacteremia (> 48–72 hours after initiation of therapy)
• Community onset
• Fever persisting more than 72 hours
• Skin findings suggesting systemic infection.

THE ROLE OF ECHOCARDIOGRAPHY

Infective endocarditis may be difficult to detect in patients with S aureus bacteremia; experts recommend routine use of echocardiography in this process.^7,8 Transesophageal echocardiography (TEE) detects more cases of endocarditis than transthoracic echocardiography (TTE),^9,10 but access, cost, and risks lead to questions about its utility.

Guidance for the use of echocardiography in S aureus bacteremia^1,10–14 continues to evolve. Consensus seems to be emerging that the risk of endocarditis is lower in patients with S aureus bacteremia who:

• Do not have a prosthetic valve or other permanent intracardiac device
• Have sterile blood cultures within 96 hours after the initial set
• Are not hemodialysis-dependent
• Developed the bacteremia in a healthcare setting
• Have no secondary focus of infection
• Have no clinical signs of infective endocarditis.

Heriot et al¹⁴ point out that studies of risk-stratification approaches to echocardiography in patients with S aureus bacteremia are difficult to interpret, as there are questions regarding the validity of the studies and the balance of the risks and benefits.¹ The question of timing of TEE remains largely unexplored, both in initial screening and in follow-up of previously undiagnosed cases of S aureus endocarditis.

In this issue of the Journal, Mirrakhimov et al¹⁵ weigh in on use of a risk-stratification model to guide use of TEE in patients with S aureus bacteremia. Their comments about avoiding TEE in patients who have an alternative explanation for S aureus bacteremia and a low pretest probability for infectious endocarditis and in patients with a disease focus that requires extended treatment are derived from a survey of infectious disease physicians.¹⁶

ROLE OF INFECTIOUS DISEASE CONSULTATION

Infectious disease consultation reduces mortality rates and healthcare costs for a variety of infections, with endocarditis as a prime example.¹⁷ For S aureus bacteremia, a large and growing body of literature demonstrates the impact of infectious disease consultation, including improved adherence to guidelines and quality measures,^18–20 lower in-hospital mortality rates^18–21 and earlier hospital discharge.¹⁸ In the era of “curbside consults” and “e-consultation,” it is interesting to note the enduring value of bedside, in-person consultation in the management of S aureus bacteremia.²⁰

Many people with S aureus bacteremia should undergo TEE. Until the evidence becomes more robust, the decision to forgo TEE must be made with caution. The expertise of infectious disease physicians in the diagnosis and management of endocarditis can assist clinicians working with the often-complex patients who develop S aureus bacteremia. If the goal is to improve outcomes, infectious disease consultation may be at least as important as appropriate selection of patients for TEE.

Morbidity and mortality rates in patients with Staphylococcus aureus bacteremia remain high even though diagnostic tests have improved and antibiotic therapy is effective. Diagnosis and management are made more complex by difficulties in finding the source of bacteremia and sites of metastatic infection.

See related article

S aureus bacteremia is a finding that demands further investigation, since up to 25% of people who have it may have endocarditis, a condition with even worse consequences.¹ The ability of S aureus to infect normal valves^2,3 adds to the challenge. In the mid-20th century, Wilson and Hamburger⁴ demonstrated that 64% of patients with S aureus bacteremia had evidence of valvular infection at autopsy. In a more recent case series of patients with S aureus endocarditis, the diagnosis was established at autopsy in 32%.⁵

Specific clinical findings in patients with complicated S aureus bacteremia—those who have a site of infection remote from or extended beyond the primary focus—may be useful in determining the need for additional diagnostic and therapeutic measures.

In a prospective cohort study, Fowler et al⁶ identified several factors that predicted complicated S aureus bacteremia (including but not limited to endocarditis):

• Prolonged bacteremia (> 48–72 hours after initiation of therapy)
• Community onset
• Fever persisting more than 72 hours
• Skin findings suggesting systemic infection.

THE ROLE OF ECHOCARDIOGRAPHY

Infective endocarditis may be difficult to detect in patients with S aureus bacteremia; experts recommend routine use of echocardiography in this process.^7,8 Transesophageal echocardiography (TEE) detects more cases of endocarditis than transthoracic echocardiography (TTE),^9,10 but access, cost, and risks lead to questions about its utility.

Guidance for the use of echocardiography in S aureus bacteremia^1,10–14 continues to evolve. Consensus seems to be emerging that the risk of endocarditis is lower in patients with S aureus bacteremia who:

• Do not have a prosthetic valve or other permanent intracardiac device
• Have sterile blood cultures within 96 hours after the initial set
• Are not hemodialysis-dependent
• Developed the bacteremia in a healthcare setting
• Have no secondary focus of infection
• Have no clinical signs of infective endocarditis.

Heriot et al¹⁴ point out that studies of risk-stratification approaches to echocardiography in patients with S aureus bacteremia are difficult to interpret, as there are questions regarding the validity of the studies and the balance of the risks and benefits.¹ The question of timing of TEE remains largely unexplored, both in initial screening and in follow-up of previously undiagnosed cases of S aureus endocarditis.

In this issue of the Journal, Mirrakhimov et al¹⁵ weigh in on use of a risk-stratification model to guide use of TEE in patients with S aureus bacteremia. Their comments about avoiding TEE in patients who have an alternative explanation for S aureus bacteremia and a low pretest probability for infectious endocarditis and in patients with a disease focus that requires extended treatment are derived from a survey of infectious disease physicians.¹⁶

ROLE OF INFECTIOUS DISEASE CONSULTATION

Infectious disease consultation reduces mortality rates and healthcare costs for a variety of infections, with endocarditis as a prime example.¹⁷ For S aureus bacteremia, a large and growing body of literature demonstrates the impact of infectious disease consultation, including improved adherence to guidelines and quality measures,^18–20 lower in-hospital mortality rates^18–21 and earlier hospital discharge.¹⁸ In the era of “curbside consults” and “e-consultation,” it is interesting to note the enduring value of bedside, in-person consultation in the management of S aureus bacteremia.²⁰

Many people with S aureus bacteremia should undergo TEE. Until the evidence becomes more robust, the decision to forgo TEE must be made with caution. The expertise of infectious disease physicians in the diagnosis and management of endocarditis can assist clinicians working with the often-complex patients who develop S aureus bacteremia. If the goal is to improve outcomes, infectious disease consultation may be at least as important as appropriate selection of patients for TEE.

Patients who survive critical illness such as shock or respiratory failure warranting admission to an intensive care unit (ICU) often develop a constellation of chronic symptoms including cognitive decline, psychiatric disturbances, and physical weakness. These changes can prevent patients from returning to their former level of function and often necessitate significant support for patients and their caregivers.¹

See related editorial

With growing awareness of the unique needs of ICU survivors, multidisciplinary PICS clinics have emerged. However, access to these clinics is limited, and most patients discharged from the ICU eventually follow up with their primary care provider. Primary care physicians who recognize PICS, understand its prognosis and its burden on caregivers, and are aware of tools that have shown promise in its management will be well prepared to address the needs of these patients.

COGNITIVE DECLINE

Several studies have shown that survivors of critical illness suffer from long-term impairment of multiple domains of cognition, including executive function. In one study, 40% of ICU survivors had global cognition scores at 1 year after discharge that were worse than those seen in moderate traumatic brain injury, and over 25% had scores similar to those seen in Alzheimer dementia.² Age had poor correlation with the incidence of long-term cognitive impairment. Cognitive impairment may not be recognized in younger patients without a high index of suspicion and directed cognitive screening. Well-known cognitive impairment screening tests such as the Montreal Cognitive Assessment may help in the evaluation of PICS.

No treatment has been shown to improve long-term cognitive impairment from any cause. The most important intervention is to recognize it and to consider how impaired executive function may interfere with other aspects of treatment, such as participation in physical therapy and adherence to medication regimens.

Evidence is also emerging that patients are often inappropriately discharged on psychoactive medications (including atypical antipsychotic drugs and sedatives) that were started in the inpatient setting.⁷ These medications increase the risk of accidents, arrhythmia, and infection, as well as add to the overall cost of postdischarge care, and they do not improve the prolonged confusion and cognitive impairment associated with PICS.⁸ Psychoactive medications should be discontinued once delirium-associated behavior has resolved, as recommended in the American Geriatrics Society guideline on postoperative delirium.⁹ Further, patients and caregivers should be counseled so that they have reasonable expectations regarding the timing of cognitive recovery, which may be prolonged and incomplete.

PHYSICAL WEAKNESS

Prolonged physical weakness may affect up to one-third of patients who survive critical illness, and it may persist for years, severely compromising quality of life.¹⁰ In addition to deconditioning due to bedrest and illness, ICU patients often develop critical illness myopathy and critical illness polyneuropathy.

Although the mechanisms and risk factors for injury to muscles and peripheral nerves are  not completely understood, the severity has been well described and ranges from proximal muscle weakness to complete quadriparesis, with inability to wean from mechanical ventilation. There is also an association with the severity of sepsis and the use of glucocorticoids and paralytics.¹⁰

Physical weakness can be readily apparent on routine history and physical examination. Differentiating critical illness myopathy from critical illness polyneuropathy requires invasive testing, including electromyography, but the results may not change management in the outpatient setting, making it unnecessary for most patients.

Physical weakness places a heavy burden on patients and their family and caregivers. As a result, most ICU patients suffer loss of employment and require supportive services on discharge, including home health aides and even institutionalization.

Physical therapy and occupational therapy are effective in reducing weakness and improving physical functioning; starting physical therapy in the outpatient setting may be as effective as early intervention in the ICU.¹¹ Given the high prevalence of respiratory and cardiovascular disease in patients after ICU discharge, referral for pulmonary or cardiovascular rehabilitation is recommended. Because of the possible link between glucocorticoids and critical illness myopathy, these drugs should be decreased or discontinued as soon as possible.

Mental health impairments in ICU survivors are common, severe, debilitating, and unfortunately, commonly overlooked. A recent study found a 37% incidence of depression and a 40% incidence of anxiety; further, 22% of patients met criteria for posttraumatic stress disorder.¹² Patients with critical illness are also more likely to have had untreated mental health illness before hospitalization. Anxiety may present with poor sleep, irritability, and fatigue. Posttraumatic stress disorder may manifest as flashbacks or as a severe cognitive or behavioral response to provocation. All of these may be assessed using standard screening questionnaires, including the Posttraumatic Stress Disorder Checklist, the 2-item Patient Health Questionnaire (PHQ-2) for depression, and the 7-item Generalized Anxiety Disorder Screen (GAD-7).

Many primary care physicians are comfortable treating some of the psychiatric disturbances associated with PICS, such as depression, but may be challenged by the spectrum and complexity of mental illness of ICU survivors. Early referral to a mental health professional ensures optimal psychiatric care and allows more time to focus on the patient’s medical comorbidities.

SOCIAL SUPPORT

The cognitive, physical, and mental health complications coupled with other medical and psychiatric comorbidities result in serious social and financial stress on patients and their families. Long-term follow-up studies show that only half of patients return to work within 1 year of critical illness and that nearly one-fourth require continued assistance with activities of daily living.¹³ Reassuringly, however, most patients in 1 study had returned to work by 2 years from discharge.³

The immense burden on caregivers, the decrease in income, and increased expenditures in providing care result in increased stress on families. The incidence of depression, anxiety, and posttraumatic stress disorder is similar among patients and their caregivers.¹¹ The frequency of emotional morbidity and the severity of the caregiver burden associated with caring for ICU survivors led to the description of a new entity: post-intensive care syndrome-family, or PICS-F.

Because of these stresses, patients often benefit from referral to a social worker. Patients should also be encouraged to bring their caregivers to physician appointments, and family members should be encouraged to discuss their perspectives in the context of a dedicated appointment. Family members should also be screened and treated for their own medical and mental health challenges. A dedicated ICU survivorship clinic may help facilitate this holistic approach and provide complementary services to the primary care provider.

CRITICAL CARE RECOVERY

As survival rates after critical illness continue to improve and clinicians encounter more patients with PICS, it is essential to appreciate the extent of associated physical, emotional, and financial hardship and to recognize when cognitive impairment may interfere with treatment. Early and accurate recognition of these challenges can help the primary care physician arrange and coordinate recovery services that ICU survivors require. Including family members in follow-up appointments can help overcome challenges in adherence to treatment plans, uncover gaps in social support, and identify signs of caregiver distress.

A thorough physical assessment and a thoughtful reconciliation of medications are critical, as is engaging the assistance of physical and occupational therapists, mental health professionals, and social workers.

Risk factors for the illness that necessitated the ICU stay such as uncontrolled diabetes, chronic obstructive pulmonary disease, and substance abuse, as well as medical sequelae such as chronic respiratory failure and heart failure, must be considered and addressed by the primary care physician, with referral to medical specialists if necessary.

Referral to an ICU survivorship center, if locally available, could help the physician manage the patient’s complex and multidisciplinary physical and neuropsychiatric needs. The Society of Critical Care Medicine maintains a resource for survivors and families at www.myicucare.org/thrive/pages/find-in-person-support-groups.aspx.

Patients who survive critical illness such as shock or respiratory failure warranting admission to an intensive care unit (ICU) often develop a constellation of chronic symptoms including cognitive decline, psychiatric disturbances, and physical weakness. These changes can prevent patients from returning to their former level of function and often necessitate significant support for patients and their caregivers.¹

See related editorial

With growing awareness of the unique needs of ICU survivors, multidisciplinary PICS clinics have emerged. However, access to these clinics is limited, and most patients discharged from the ICU eventually follow up with their primary care provider. Primary care physicians who recognize PICS, understand its prognosis and its burden on caregivers, and are aware of tools that have shown promise in its management will be well prepared to address the needs of these patients.

COGNITIVE DECLINE

Several studies have shown that survivors of critical illness suffer from long-term impairment of multiple domains of cognition, including executive function. In one study, 40% of ICU survivors had global cognition scores at 1 year after discharge that were worse than those seen in moderate traumatic brain injury, and over 25% had scores similar to those seen in Alzheimer dementia.² Age had poor correlation with the incidence of long-term cognitive impairment. Cognitive impairment may not be recognized in younger patients without a high index of suspicion and directed cognitive screening. Well-known cognitive impairment screening tests such as the Montreal Cognitive Assessment may help in the evaluation of PICS.

No treatment has been shown to improve long-term cognitive impairment from any cause. The most important intervention is to recognize it and to consider how impaired executive function may interfere with other aspects of treatment, such as participation in physical therapy and adherence to medication regimens.

Evidence is also emerging that patients are often inappropriately discharged on psychoactive medications (including atypical antipsychotic drugs and sedatives) that were started in the inpatient setting.⁷ These medications increase the risk of accidents, arrhythmia, and infection, as well as add to the overall cost of postdischarge care, and they do not improve the prolonged confusion and cognitive impairment associated with PICS.⁸ Psychoactive medications should be discontinued once delirium-associated behavior has resolved, as recommended in the American Geriatrics Society guideline on postoperative delirium.⁹ Further, patients and caregivers should be counseled so that they have reasonable expectations regarding the timing of cognitive recovery, which may be prolonged and incomplete.

PHYSICAL WEAKNESS

Prolonged physical weakness may affect up to one-third of patients who survive critical illness, and it may persist for years, severely compromising quality of life.¹⁰ In addition to deconditioning due to bedrest and illness, ICU patients often develop critical illness myopathy and critical illness polyneuropathy.

Although the mechanisms and risk factors for injury to muscles and peripheral nerves are  not completely understood, the severity has been well described and ranges from proximal muscle weakness to complete quadriparesis, with inability to wean from mechanical ventilation. There is also an association with the severity of sepsis and the use of glucocorticoids and paralytics.¹⁰

Physical weakness can be readily apparent on routine history and physical examination. Differentiating critical illness myopathy from critical illness polyneuropathy requires invasive testing, including electromyography, but the results may not change management in the outpatient setting, making it unnecessary for most patients.

Physical weakness places a heavy burden on patients and their family and caregivers. As a result, most ICU patients suffer loss of employment and require supportive services on discharge, including home health aides and even institutionalization.

Physical therapy and occupational therapy are effective in reducing weakness and improving physical functioning; starting physical therapy in the outpatient setting may be as effective as early intervention in the ICU.¹¹ Given the high prevalence of respiratory and cardiovascular disease in patients after ICU discharge, referral for pulmonary or cardiovascular rehabilitation is recommended. Because of the possible link between glucocorticoids and critical illness myopathy, these drugs should be decreased or discontinued as soon as possible.

Mental health impairments in ICU survivors are common, severe, debilitating, and unfortunately, commonly overlooked. A recent study found a 37% incidence of depression and a 40% incidence of anxiety; further, 22% of patients met criteria for posttraumatic stress disorder.¹² Patients with critical illness are also more likely to have had untreated mental health illness before hospitalization. Anxiety may present with poor sleep, irritability, and fatigue. Posttraumatic stress disorder may manifest as flashbacks or as a severe cognitive or behavioral response to provocation. All of these may be assessed using standard screening questionnaires, including the Posttraumatic Stress Disorder Checklist, the 2-item Patient Health Questionnaire (PHQ-2) for depression, and the 7-item Generalized Anxiety Disorder Screen (GAD-7).

Many primary care physicians are comfortable treating some of the psychiatric disturbances associated with PICS, such as depression, but may be challenged by the spectrum and complexity of mental illness of ICU survivors. Early referral to a mental health professional ensures optimal psychiatric care and allows more time to focus on the patient’s medical comorbidities.

SOCIAL SUPPORT

The cognitive, physical, and mental health complications coupled with other medical and psychiatric comorbidities result in serious social and financial stress on patients and their families. Long-term follow-up studies show that only half of patients return to work within 1 year of critical illness and that nearly one-fourth require continued assistance with activities of daily living.¹³ Reassuringly, however, most patients in 1 study had returned to work by 2 years from discharge.³

The immense burden on caregivers, the decrease in income, and increased expenditures in providing care result in increased stress on families. The incidence of depression, anxiety, and posttraumatic stress disorder is similar among patients and their caregivers.¹¹ The frequency of emotional morbidity and the severity of the caregiver burden associated with caring for ICU survivors led to the description of a new entity: post-intensive care syndrome-family, or PICS-F.

Because of these stresses, patients often benefit from referral to a social worker. Patients should also be encouraged to bring their caregivers to physician appointments, and family members should be encouraged to discuss their perspectives in the context of a dedicated appointment. Family members should also be screened and treated for their own medical and mental health challenges. A dedicated ICU survivorship clinic may help facilitate this holistic approach and provide complementary services to the primary care provider.

CRITICAL CARE RECOVERY

As survival rates after critical illness continue to improve and clinicians encounter more patients with PICS, it is essential to appreciate the extent of associated physical, emotional, and financial hardship and to recognize when cognitive impairment may interfere with treatment. Early and accurate recognition of these challenges can help the primary care physician arrange and coordinate recovery services that ICU survivors require. Including family members in follow-up appointments can help overcome challenges in adherence to treatment plans, uncover gaps in social support, and identify signs of caregiver distress.

A thorough physical assessment and a thoughtful reconciliation of medications are critical, as is engaging the assistance of physical and occupational therapists, mental health professionals, and social workers.

Risk factors for the illness that necessitated the ICU stay such as uncontrolled diabetes, chronic obstructive pulmonary disease, and substance abuse, as well as medical sequelae such as chronic respiratory failure and heart failure, must be considered and addressed by the primary care physician, with referral to medical specialists if necessary.

Referral to an ICU survivorship center, if locally available, could help the physician manage the patient’s complex and multidisciplinary physical and neuropsychiatric needs. The Society of Critical Care Medicine maintains a resource for survivors and families at www.myicucare.org/thrive/pages/find-in-person-support-groups.aspx.

Patients who survive critical illness such as shock or respiratory failure warranting admission to an intensive care unit (ICU) often develop a constellation of chronic symptoms including cognitive decline, psychiatric disturbances, and physical weakness. These changes can prevent patients from returning to their former level of function and often necessitate significant support for patients and their caregivers.¹

See related editorial

With growing awareness of the unique needs of ICU survivors, multidisciplinary PICS clinics have emerged. However, access to these clinics is limited, and most patients discharged from the ICU eventually follow up with their primary care provider. Primary care physicians who recognize PICS, understand its prognosis and its burden on caregivers, and are aware of tools that have shown promise in its management will be well prepared to address the needs of these patients.

COGNITIVE DECLINE

Several studies have shown that survivors of critical illness suffer from long-term impairment of multiple domains of cognition, including executive function. In one study, 40% of ICU survivors had global cognition scores at 1 year after discharge that were worse than those seen in moderate traumatic brain injury, and over 25% had scores similar to those seen in Alzheimer dementia.² Age had poor correlation with the incidence of long-term cognitive impairment. Cognitive impairment may not be recognized in younger patients without a high index of suspicion and directed cognitive screening. Well-known cognitive impairment screening tests such as the Montreal Cognitive Assessment may help in the evaluation of PICS.

No treatment has been shown to improve long-term cognitive impairment from any cause. The most important intervention is to recognize it and to consider how impaired executive function may interfere with other aspects of treatment, such as participation in physical therapy and adherence to medication regimens.

Evidence is also emerging that patients are often inappropriately discharged on psychoactive medications (including atypical antipsychotic drugs and sedatives) that were started in the inpatient setting.⁷ These medications increase the risk of accidents, arrhythmia, and infection, as well as add to the overall cost of postdischarge care, and they do not improve the prolonged confusion and cognitive impairment associated with PICS.⁸ Psychoactive medications should be discontinued once delirium-associated behavior has resolved, as recommended in the American Geriatrics Society guideline on postoperative delirium.⁹ Further, patients and caregivers should be counseled so that they have reasonable expectations regarding the timing of cognitive recovery, which may be prolonged and incomplete.

PHYSICAL WEAKNESS

Prolonged physical weakness may affect up to one-third of patients who survive critical illness, and it may persist for years, severely compromising quality of life.¹⁰ In addition to deconditioning due to bedrest and illness, ICU patients often develop critical illness myopathy and critical illness polyneuropathy.

Although the mechanisms and risk factors for injury to muscles and peripheral nerves are  not completely understood, the severity has been well described and ranges from proximal muscle weakness to complete quadriparesis, with inability to wean from mechanical ventilation. There is also an association with the severity of sepsis and the use of glucocorticoids and paralytics.¹⁰

Physical weakness can be readily apparent on routine history and physical examination. Differentiating critical illness myopathy from critical illness polyneuropathy requires invasive testing, including electromyography, but the results may not change management in the outpatient setting, making it unnecessary for most patients.

Physical weakness places a heavy burden on patients and their family and caregivers. As a result, most ICU patients suffer loss of employment and require supportive services on discharge, including home health aides and even institutionalization.

Physical therapy and occupational therapy are effective in reducing weakness and improving physical functioning; starting physical therapy in the outpatient setting may be as effective as early intervention in the ICU.¹¹ Given the high prevalence of respiratory and cardiovascular disease in patients after ICU discharge, referral for pulmonary or cardiovascular rehabilitation is recommended. Because of the possible link between glucocorticoids and critical illness myopathy, these drugs should be decreased or discontinued as soon as possible.

Mental health impairments in ICU survivors are common, severe, debilitating, and unfortunately, commonly overlooked. A recent study found a 37% incidence of depression and a 40% incidence of anxiety; further, 22% of patients met criteria for posttraumatic stress disorder.¹² Patients with critical illness are also more likely to have had untreated mental health illness before hospitalization. Anxiety may present with poor sleep, irritability, and fatigue. Posttraumatic stress disorder may manifest as flashbacks or as a severe cognitive or behavioral response to provocation. All of these may be assessed using standard screening questionnaires, including the Posttraumatic Stress Disorder Checklist, the 2-item Patient Health Questionnaire (PHQ-2) for depression, and the 7-item Generalized Anxiety Disorder Screen (GAD-7).

Many primary care physicians are comfortable treating some of the psychiatric disturbances associated with PICS, such as depression, but may be challenged by the spectrum and complexity of mental illness of ICU survivors. Early referral to a mental health professional ensures optimal psychiatric care and allows more time to focus on the patient’s medical comorbidities.

SOCIAL SUPPORT

The cognitive, physical, and mental health complications coupled with other medical and psychiatric comorbidities result in serious social and financial stress on patients and their families. Long-term follow-up studies show that only half of patients return to work within 1 year of critical illness and that nearly one-fourth require continued assistance with activities of daily living.¹³ Reassuringly, however, most patients in 1 study had returned to work by 2 years from discharge.³

The immense burden on caregivers, the decrease in income, and increased expenditures in providing care result in increased stress on families. The incidence of depression, anxiety, and posttraumatic stress disorder is similar among patients and their caregivers.¹¹ The frequency of emotional morbidity and the severity of the caregiver burden associated with caring for ICU survivors led to the description of a new entity: post-intensive care syndrome-family, or PICS-F.

Because of these stresses, patients often benefit from referral to a social worker. Patients should also be encouraged to bring their caregivers to physician appointments, and family members should be encouraged to discuss their perspectives in the context of a dedicated appointment. Family members should also be screened and treated for their own medical and mental health challenges. A dedicated ICU survivorship clinic may help facilitate this holistic approach and provide complementary services to the primary care provider.

CRITICAL CARE RECOVERY

As survival rates after critical illness continue to improve and clinicians encounter more patients with PICS, it is essential to appreciate the extent of associated physical, emotional, and financial hardship and to recognize when cognitive impairment may interfere with treatment. Early and accurate recognition of these challenges can help the primary care physician arrange and coordinate recovery services that ICU survivors require. Including family members in follow-up appointments can help overcome challenges in adherence to treatment plans, uncover gaps in social support, and identify signs of caregiver distress.

A thorough physical assessment and a thoughtful reconciliation of medications are critical, as is engaging the assistance of physical and occupational therapists, mental health professionals, and social workers.

Risk factors for the illness that necessitated the ICU stay such as uncontrolled diabetes, chronic obstructive pulmonary disease, and substance abuse, as well as medical sequelae such as chronic respiratory failure and heart failure, must be considered and addressed by the primary care physician, with referral to medical specialists if necessary.

Referral to an ICU survivorship center, if locally available, could help the physician manage the patient’s complex and multidisciplinary physical and neuropsychiatric needs. The Society of Critical Care Medicine maintains a resource for survivors and families at www.myicucare.org/thrive/pages/find-in-person-support-groups.aspx.

My introduction to critical care medicine came about during the summer between my third and fourth years of medical school. During that brief break, I, like most of my classmates, was drawn to the classic medical satire The House of God by Samuel Shem,¹ which had become a cult classic in the medical field for its ghoulish medical wisdom and dark humor. In “the house,” the intensive care unit (ICU) is “that mausoleum down the hall,” its patients “perched precariously on the edge of that slick bobsled ride down to death.”¹ This sentiment persisted even as I began my critical care medicine fellowship in the mid-1990s.

See related article

The science and practice of critical care medicine have changed, evolved, and advanced over the past several decades reflecting newer technology, but also an aging population with higher acuity.² Critical care medicine has established itself as a specialty in its own right, and the importance of the physician intensivist-led multidisciplinary care teams in optimizing outcome has been demonstrated.^3,4 These teams have been associated with improved quality of care, reduced length of stay, improved resource utilization, and reduced rates of complications, morbidity, and death.

While there have been few medical miracles and limited advances in therapeutics over the last 30 years, advances in patient management, adherence to processes of care, better use of technology, and more timely diagnosis and treatment have facilitated improved outcomes.⁵ Collaboration with nurses, respiratory therapists, pharmacists, and other healthcare personnel is invaluable, as these providers are responsible for executing management protocols such as weaning sedation and mechanical ventilation, nutrition, glucose control, vasopressor and electrolyte titration, positioning, and early ambulation.

Unfortunately, as an increasing number of patients are being discharged from the ICU, evidence is accumulating that ICU survivors may develop persistent organ dysfunction requiring prolonged stays in the ICU and resulting in chronic critical illness. A 2015 study estimated 380,000 cases of chronic critical illness annually, particularly among the elderly population, with attendant hospital costs of up to $26 billion.⁶ While 70% of these patients may survive their hospitalization, the Society of Critical Care Medicine (SCCM) estimates that the 1-year post-discharge mortality rate may exceed 50%.⁷

We can take pride and comfort in knowing that the past several decades have seen growth in critical care training, more engaged practice, and heightened communication resulting in lower mortality rates.⁸ However, a majority of survivors suffer significant morbidities that may be severe and persist for a prolonged period after hospital discharge. These worsening impairments after discharge are termed postintensive care syndrome (PICS), which manifests as a new or worsening mental, cognitive, and physical condition and may affect up to 50% of ICU survivors.⁶

The impact on daily functioning and quality of life can be devastating, and primary care physicians will be increasingly called on to diagnose and participate in ongoing post-discharge management. Additionally, the impact of critical illness on relatives and informal caregivers can be long-lasting and profound, increasing their own risk of depression, posttraumatic stress disorder, and financial hardship.

In this issue of the Journal, Golovyan and colleagues identify several potential complications and sequelae of critical illness after discharge from the ICU.⁹ Primary care providers will see these patients in outpatient settings and need to be prepared to triage and treat the new-onset and chronic conditions for which these patients are at high risk.

In addition, as the authors point out, family members and informal caregivers need to be counseled about the proper care of these patients as well as themselves.

The current healthcare system does not appropriately address these survivors and their families. In 2015, the Society of Critical Care Medicine announced the THRIVE initiative, designed to improve support for the patient and family after critical illness. Given the many survivors and caregivers touched by critical illness, the Society has invested in THRIVE with the intent of helping those affected to work together with clinicians to advance recovery. Through peer support groups, post-ICU clinics, and continuing research into quality improvement, THRIVE may help to reduce readmissions and improve quality of life for critical care survivors and their loved ones.

Things have changed since the days of The House of God. Critical care medicine has become a vibrant medical specialty and an integral part of our healthcare system. Dedicated critical care physicians and the multidisciplinary teams they lead have improved outcomes and resource utilization.^2–5

The demand for ICU care will continue to increase as our population ages and the need for medical and surgical services increases commensurately. The ratio of ICU beds to hospital beds continues to escalate, and it is feared that the demand for critical care professionals may outstrip the supply.

While we no longer see that mournful shaking of the head when a patient is admitted to the ICU, we need to have the proper vision and use the most up-to-date scientific knowledge and research in treating underlying illness to ensure that once these patients are discharged, communication continues between critical care and primary care providers. This ongoing support will ensure these patients the best possible quality of life.

My introduction to critical care medicine came about during the summer between my third and fourth years of medical school. During that brief break, I, like most of my classmates, was drawn to the classic medical satire The House of God by Samuel Shem,¹ which had become a cult classic in the medical field for its ghoulish medical wisdom and dark humor. In “the house,” the intensive care unit (ICU) is “that mausoleum down the hall,” its patients “perched precariously on the edge of that slick bobsled ride down to death.”¹ This sentiment persisted even as I began my critical care medicine fellowship in the mid-1990s.

See related article

The science and practice of critical care medicine have changed, evolved, and advanced over the past several decades reflecting newer technology, but also an aging population with higher acuity.² Critical care medicine has established itself as a specialty in its own right, and the importance of the physician intensivist-led multidisciplinary care teams in optimizing outcome has been demonstrated.^3,4 These teams have been associated with improved quality of care, reduced length of stay, improved resource utilization, and reduced rates of complications, morbidity, and death.

While there have been few medical miracles and limited advances in therapeutics over the last 30 years, advances in patient management, adherence to processes of care, better use of technology, and more timely diagnosis and treatment have facilitated improved outcomes.⁵ Collaboration with nurses, respiratory therapists, pharmacists, and other healthcare personnel is invaluable, as these providers are responsible for executing management protocols such as weaning sedation and mechanical ventilation, nutrition, glucose control, vasopressor and electrolyte titration, positioning, and early ambulation.

Unfortunately, as an increasing number of patients are being discharged from the ICU, evidence is accumulating that ICU survivors may develop persistent organ dysfunction requiring prolonged stays in the ICU and resulting in chronic critical illness. A 2015 study estimated 380,000 cases of chronic critical illness annually, particularly among the elderly population, with attendant hospital costs of up to $26 billion.⁶ While 70% of these patients may survive their hospitalization, the Society of Critical Care Medicine (SCCM) estimates that the 1-year post-discharge mortality rate may exceed 50%.⁷

We can take pride and comfort in knowing that the past several decades have seen growth in critical care training, more engaged practice, and heightened communication resulting in lower mortality rates.⁸ However, a majority of survivors suffer significant morbidities that may be severe and persist for a prolonged period after hospital discharge. These worsening impairments after discharge are termed postintensive care syndrome (PICS), which manifests as a new or worsening mental, cognitive, and physical condition and may affect up to 50% of ICU survivors.⁶

The impact on daily functioning and quality of life can be devastating, and primary care physicians will be increasingly called on to diagnose and participate in ongoing post-discharge management. Additionally, the impact of critical illness on relatives and informal caregivers can be long-lasting and profound, increasing their own risk of depression, posttraumatic stress disorder, and financial hardship.

In this issue of the Journal, Golovyan and colleagues identify several potential complications and sequelae of critical illness after discharge from the ICU.⁹ Primary care providers will see these patients in outpatient settings and need to be prepared to triage and treat the new-onset and chronic conditions for which these patients are at high risk.

In addition, as the authors point out, family members and informal caregivers need to be counseled about the proper care of these patients as well as themselves.

The current healthcare system does not appropriately address these survivors and their families. In 2015, the Society of Critical Care Medicine announced the THRIVE initiative, designed to improve support for the patient and family after critical illness. Given the many survivors and caregivers touched by critical illness, the Society has invested in THRIVE with the intent of helping those affected to work together with clinicians to advance recovery. Through peer support groups, post-ICU clinics, and continuing research into quality improvement, THRIVE may help to reduce readmissions and improve quality of life for critical care survivors and their loved ones.

Things have changed since the days of The House of God. Critical care medicine has become a vibrant medical specialty and an integral part of our healthcare system. Dedicated critical care physicians and the multidisciplinary teams they lead have improved outcomes and resource utilization.^2–5

The demand for ICU care will continue to increase as our population ages and the need for medical and surgical services increases commensurately. The ratio of ICU beds to hospital beds continues to escalate, and it is feared that the demand for critical care professionals may outstrip the supply.

While we no longer see that mournful shaking of the head when a patient is admitted to the ICU, we need to have the proper vision and use the most up-to-date scientific knowledge and research in treating underlying illness to ensure that once these patients are discharged, communication continues between critical care and primary care providers. This ongoing support will ensure these patients the best possible quality of life.

My introduction to critical care medicine came about during the summer between my third and fourth years of medical school. During that brief break, I, like most of my classmates, was drawn to the classic medical satire The House of God by Samuel Shem,¹ which had become a cult classic in the medical field for its ghoulish medical wisdom and dark humor. In “the house,” the intensive care unit (ICU) is “that mausoleum down the hall,” its patients “perched precariously on the edge of that slick bobsled ride down to death.”¹ This sentiment persisted even as I began my critical care medicine fellowship in the mid-1990s.

See related article

The science and practice of critical care medicine have changed, evolved, and advanced over the past several decades reflecting newer technology, but also an aging population with higher acuity.² Critical care medicine has established itself as a specialty in its own right, and the importance of the physician intensivist-led multidisciplinary care teams in optimizing outcome has been demonstrated.^3,4 These teams have been associated with improved quality of care, reduced length of stay, improved resource utilization, and reduced rates of complications, morbidity, and death.

While there have been few medical miracles and limited advances in therapeutics over the last 30 years, advances in patient management, adherence to processes of care, better use of technology, and more timely diagnosis and treatment have facilitated improved outcomes.⁵ Collaboration with nurses, respiratory therapists, pharmacists, and other healthcare personnel is invaluable, as these providers are responsible for executing management protocols such as weaning sedation and mechanical ventilation, nutrition, glucose control, vasopressor and electrolyte titration, positioning, and early ambulation.

Unfortunately, as an increasing number of patients are being discharged from the ICU, evidence is accumulating that ICU survivors may develop persistent organ dysfunction requiring prolonged stays in the ICU and resulting in chronic critical illness. A 2015 study estimated 380,000 cases of chronic critical illness annually, particularly among the elderly population, with attendant hospital costs of up to $26 billion.⁶ While 70% of these patients may survive their hospitalization, the Society of Critical Care Medicine (SCCM) estimates that the 1-year post-discharge mortality rate may exceed 50%.⁷

We can take pride and comfort in knowing that the past several decades have seen growth in critical care training, more engaged practice, and heightened communication resulting in lower mortality rates.⁸ However, a majority of survivors suffer significant morbidities that may be severe and persist for a prolonged period after hospital discharge. These worsening impairments after discharge are termed postintensive care syndrome (PICS), which manifests as a new or worsening mental, cognitive, and physical condition and may affect up to 50% of ICU survivors.⁶

The impact on daily functioning and quality of life can be devastating, and primary care physicians will be increasingly called on to diagnose and participate in ongoing post-discharge management. Additionally, the impact of critical illness on relatives and informal caregivers can be long-lasting and profound, increasing their own risk of depression, posttraumatic stress disorder, and financial hardship.

In this issue of the Journal, Golovyan and colleagues identify several potential complications and sequelae of critical illness after discharge from the ICU.⁹ Primary care providers will see these patients in outpatient settings and need to be prepared to triage and treat the new-onset and chronic conditions for which these patients are at high risk.

In addition, as the authors point out, family members and informal caregivers need to be counseled about the proper care of these patients as well as themselves.

The current healthcare system does not appropriately address these survivors and their families. In 2015, the Society of Critical Care Medicine announced the THRIVE initiative, designed to improve support for the patient and family after critical illness. Given the many survivors and caregivers touched by critical illness, the Society has invested in THRIVE with the intent of helping those affected to work together with clinicians to advance recovery. Through peer support groups, post-ICU clinics, and continuing research into quality improvement, THRIVE may help to reduce readmissions and improve quality of life for critical care survivors and their loved ones.

Things have changed since the days of The House of God. Critical care medicine has become a vibrant medical specialty and an integral part of our healthcare system. Dedicated critical care physicians and the multidisciplinary teams they lead have improved outcomes and resource utilization.^2–5

The demand for ICU care will continue to increase as our population ages and the need for medical and surgical services increases commensurately. The ratio of ICU beds to hospital beds continues to escalate, and it is feared that the demand for critical care professionals may outstrip the supply.

While we no longer see that mournful shaking of the head when a patient is admitted to the ICU, we need to have the proper vision and use the most up-to-date scientific knowledge and research in treating underlying illness to ensure that once these patients are discharged, communication continues between critical care and primary care providers. This ongoing support will ensure these patients the best possible quality of life.

Although calcium and vitamin D are often recommended for prevention and treatment of osteoporosis, considerable controversy exists in terms of their safety and efficacy.¹ This article highlights the issues, referring  readers to reviews and meta-analyses for details and providing some practical advice for patients requiring supplementation.

CALCIUM INTAKE AND BONE DENSITY

Calcium enters the body through diet and supplementation. If intake is low, blood calcium levels fall, resulting in secondary hyperparathyroidism, which has 3 main effects:

• Increased fractional absorption of the calcium that is consumed
• Reduced urinary excretion of calcium
• Increased bone resorption, which releases calcium into the blood,² which explains the potential for the deleterious effect of deficient intake of calcium on bone.³

Based on the simple physiology outlined above, it seems logical that insufficient intake of calcium over time could lead to mobilization of calcium from bone, lower bone mineral density, and higher fracture risk.³ This topic has been reviewed by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis, and Musculoskeletal Diseases and the International Foundation for Osteoporosis.¹

Many lines of evidence suggest that low calcium intake adversely affects bone mineral density.¹ Low calcium intake has been associated with lower bone density in some cross-sectional studies,^4–6 though not all.⁷ Interventions to increase calcium intake in postmenopausal women have shown beneficial effects on bone density,^8–10 though in some studies the benefit was small and nonprogressive.¹¹ The question is whether this improvement in bone mineral density translates into fewer fractures.

Results from individual studies looking at fracture prevention through calcium supplementation have been conflicting,^10,12–14 and reviews and meta-analyses have summarized the data.^1,3,15 A recent review of these meta-analyses showed a small but significant reduction in some types of fracture.¹

Some speculate that the difficulty in demonstrating fracture efficacy might be due to imperfect compliance with calcium intake, and that the participants in the placebo groups often had fairly robust calcium intake from diet and off-study supplemental intake, which could reduce the sensitivity of studies to demonstrate the fracture benefit.^1,16

The US Preventive Services Task Force¹⁷ recommends that the general public not take supplemental calcium for skeletal health, but emphasizes that this recommendation does not apply to patients with osteoporosis. Most other official guidelines (eg, those of the Endocrine Society,¹⁸ American Association of Clinical Endocrinologists,¹⁹ Institute of Medicine,²⁰ and National Osteoporosis Foundation²¹) recommend adequate calcium intake to optimize skeletal health.

CALCIUM INTAKE AND CORONARY ARTERY DISEASE

Patients often wonder if the calcium in their supplements ends up in their coronary arteries rather than their bones. Although we once dismissed such concerns, several studies and meta-analyses have reported higher rates of cardiovascular disease with supplemental calcium use.^22–24 A proposed mechanism to explain this increased risk is that taking calcium supplements transiently raises the serum calcium level, resulting in calcium deposition in coronary arteries, accelerating atherosclerosis formation.²⁵

On the other hand, some studies and meta-analyses have not shown any increased risk of cardiovascular disease with calcium and vitamin D supplementation.^26,27 This subject has been reviewed by Harvey et al.¹

Our conclusions are as follows:

Patients should be told that the National Osteoporosis Foundation and the American Society for Preventive Cardiology released a statement in 2016 adopting the position that calcium intake from food and supplements should be considered safe from a cardiovascular perspective.²⁸

If patients want to avoid the possible increase in risk of cardiovascular disease due to calcium supplementation, they can optimize their calcium intake with dietary calcium. Observational studies that showed increased risk with supplemental calcium found no such increase in cardiovascular disease with a robust dietary intake of calcium.²⁹

This is not to say that patients should be encouraged to boost their dietary calcium intake and avoid heart disease by eating more cheese and ice cream, as these foods are high in saturated fats and cholesterol. Many dairy and nondairy sources of calcium do not contain these undesirable nutrients.

CALCIUM SUPPLEMENTATION AND NEPHROLITHIASIS

High dietary calcium intake has not been shown to increase the risk of kidney stones.

In the Nurses’ Health Study, the multivariate relative risk of stone formation was 0.65 (95% confidence interval [CI] 0.5–0.83) in those in the highest vs the lowest quintiles of dietary calcium intake.³⁰ In contrast, the relative risk of stones in those taking calcium supplements was 1.2 (CI 1.02–1.41),³⁰ although this higher risk was not seen in younger women (ages 27 to 44).³¹  

Similar results were seen in the Women’s Health Initiative, in which calcium carbonate and vitamin D supplements resulted in a relative increased risk of stone formation of 1.17 (95% CI 1.02–1.34) compared with women on placebo.¹²

Data from male stone-formers also suggests that high dietary calcium intake does not increase the risk of stones.³²

A theory to explain the difference between dietary and supplemental calcium with respect to stone formation is that dietary calcium binds to oxalate in the gut and reduces its absorption. The most common type of kidney stones are composed of calcium oxalate, and the oxalate, not the calcium, may be the real culprit. In contrast, calcium supplements are often taken between meals and therefore do not exert this protective effect and may be absorbed more rapidly and raise the serum calcium level more, which could lead to higher urinary calcium excretion.³³

CALCIUM INTAKE IN PATIENTS TAKING ANTIRESORPTIVE DRUGS

Patients often mistakenly think that calcium and vitamin D supplements are given for mild cases of bone loss, and that if their bone loss is significant enough to require a medication, then they no longer need calcium and vitamin D supplements. Most clinical trials showing bone mineral density and fracture benefit from antiresorptive therapy were in patients who were taking enough calcium and vitamin D, so the efficacy of antiresorptive therapy is most clear only when taking enough calcium and vitamin D.^34,35

Furthermore, patients with inadequate calcium and vitamin D intake essentially maintain their serum calcium levels by mobilizing calcium from bone; the combination of insufficient calcium intake and administration of agents that interfere with the ability to mobilize calcium from bone may put patients at risk of hypocalcemia.³⁶

OPTIMIZING INTAKE OF CALCIUM AND VITAMIN D

Diet is key to calcium intake. People should consume adequate amounts of calcium-rich foods regardless of whether they have a history of kidney stones, since robust dietary intake of calcium does not increase the risk of cardiovascular disease or kidney stones and may actually have a protective effect. We also remain skeptical of the concern that supplemental calcium increases the risk of cardiovascular disease.

We recommend a target total calcium intake from diet, and if necessary, supplements, of 1,000 to 1,200 mg daily, and not to worry about cardiovascular disease or kidney stones. A patient or clinician reluctant to push calcium intake that high with supplements might opt for a more conservative goal of 800 mg of calcium daily. This recommendation is based on data suggesting that in the presence of vitamin D sufficiency, calcium supplementation with 500 mg of calcium citrate does little for patients whose calcium intake is above 400 mg/day.³⁷

Vitamin D: How much do we need?

Regarding vitamin D intake, the Institute of Medicine recommends 600 to 800 IU to achieve a 25-hydroxyvitamin D level of 20 to 40 ng/mL.²⁰

The Endocrine Society recommends “at least” 600 to 800 IU, but says that 1,500 to 2,000 IU may be needed to get the 25-hydroxyvitamin D level to 30 to 60 ng/mL.¹⁸

The Institute of Medicine based its recommendation on randomized controlled trials that showed fewer fractures with vitamin D intakes of 600 to 800 IU/day.^13,14 Also, observational studies show little further reduction in fracture risk when the 25-hydroxyvitamin D levels rise above 20 ng/mL.³⁸ A case-control study found an association between 25-hydroxyvitamin D levels higher than 40 ng/mL and pancreatic cancer.³⁹

The Endocrine Society guidelines recommended higher intakes and levels of vitamin D because there are data suggesting that vitamin D levels higher than 30 ng/mL suppress parathyroid hormone levels further, which should favor less mobilization of bone.⁴⁰

Levels of 25-hydroxyvitamin D in people exposed to plenty of sunlight rarely go above 60 ng/mL, suggesting 60 ng/mL should be the upper limit of levels to target, and it is unlikely that such levels are harmful.⁴¹

Implementing either recommendation—a target 25-hydroxyvitamin D level of 20 to 40 ng/mL or 30 to 60 ng/mL—is reasonable.

CALCULATING A PATIENT’S DIETARY CALCIUM INTAKE

A detailed dietary history can be obtained by a dietitian, or by using the Calcium Calculator app supported by the International Osteoporosis Foundation.⁴³ However, dietary calcium intake can be assessed quickly. To approximate a patient’s total dietary calcium intake (in milligrams), we multiply the number of servings of dietary calcium by 300. A serving of dietary calcium is found in:

• 1 cup of milk, yogurt, calcium-fortified juice, almonds, cooked spinach, or collard greens
• 1.5 ounces of hard cheese
• 2 cups of ice cream, cottage cheese, or beans
• 4 ounces of tofu or canned fish with bones such as salmon or sardines.

Therefore, if a patient consumes 1 cup of milk daily and 1 cup of yogurt 3 times a week, she takes in an estimated 1.5 servings of dietary calcium daily, or 450 mg. What the patient does not receive in the diet should be made up with supplemental calcium.

CALCIUM SUPPLEMENTS

Calcium citrate has certain advantages as a supplement. Calcium carbonate requires gastric acidity to be absorbed and is therefore better absorbed if taken with meals; however, calcium citrate is equally well absorbed in the fasting or fed state and so can be taken without regard for achlorhydria or timing of meals.⁴⁴

Another potential advantage of calcium citrate is that it has never been shown to increase the risk of kidney stones the way calcium carbonate has.¹² Further, potassium citrate is a treatment for certain types of kidney stones,⁴⁵ and it is possible that when calcium is given as citrate there is less danger of kidney stones.⁴⁶ For these reasons, we generally recommend calcium citrate over other forms of calcium.

The brand of calcium citrate most readily available is Citracal, but any version of calcium citrate is acceptable.

SOURCES OF CONFUSION

Labels that describe calcium content of supplements are often misleading, and this lack of clarity can interfere with the patient’s ability to correctly identify how much calcium is in each pill.

Serving size. Whereas 1 serving of Caltrate is 1 pill, 1 serving of Tums or Citracal is 2 pills; for other brands a serving may be 3 or 4 pills.

Calcium salt vs elemental calcium. The amount of elemental calcium contained in different calcium salts varies according to the molecular weight of the salt: 1,000 mg of calcium carbonate has 400 mg of elemental calcium, while 1,000 mg of calcium citrate has 200 mg of elemental calcium.

When we recommend 1,000 to 1,200 mg of calcium daily, we mean the amount of elemental calcium. The label on calcium supplements usually indicates the amount of elemental calcium, but some have confusing information about the amount of calcium salt they contain. For instance, Tums lists the amount of calcium carbonate per pill on the top of the label, but elsewhere lists the amount of elemental calcium.

Same brand, different preparation. Some brands of calcium have more than 1 formulation, each with a different amount of calcium. For instance, Citracal has a maximum-strength 315-mg tablet and a “petite” 200-mg tablet. Careful reading of the label is required to make sure that the patient is getting the amount of calcium she thinks she is getting.

OPTIONS FOR THOSE WITH DIFFICULTY SWALLOWING LARGE PILLS

Many calcium pills are large and difficult to swallow. Patients often ask if calcium pills can be crushed, and the answer is that they certainly can, but this approach is cumbersome and usually results in patients eventually stopping calcium in frustration.

CALCIUM SUPPLEMENTS AND CONSTIPATION

Constipation is a common side effect of calcium supplementation.⁴⁷ Many patients report that they cannot take a calcium supplement because of constipation, or ask if there are calcium preparations that are less constipating than others.

There are ways of overcoming the constipating effects of calcium. Osmotic laxatives and stool softeners such as polyethylene glycol, magnesium citrate, and docusate sodium are safe and effective, although patients are often reluctant to take a medicine to combat the side effects from another medicine.

In such circumstances patients are often amenable to taking a combination product such as calcium with magnesium, since the cathartic effects of magnesium nicely counteract the constipating effects of calcium. This idea is exploited in antacids such as Rolaids, which are combinations of calcium carbonate and magnesium oxide that usually have no net effect on stool consistency.⁴⁷

Many patients believe that calcium must be combined with magnesium to be absorbed. Although there are no data to support this idea, a patient already harboring this misconception may be more amenable to calcium-magnesium combinations for the purpose of avoiding constipation.

If a patient cannot find a calcium preparation that she can take at the full recommended doses, we often suggest starting with a very small dose for 2 weeks, and then adjusting the dose upward every 2 weeks until reaching the maximum dose that the patient can tolerate. Even if the dose is well below recommended doses, most of the benefit of calcium is obtained by bringing total intake to more than 500 mg daily,³⁷ so continued use should be encouraged even when optimal targets cannot be sustained.

For patients who cannot tolerate enough calcium, we recommend being especially sure to optimize the vitamin D levels, since there are studies that suggest that secondary hyperparathyroidism mostly occurs in states of low calcium intake if vitamin D levels are insufficient.⁴⁸

If the patient has secondary hyperpara­thyroidism despite best attempts at supplementation with calcium and vitamin D, consider prescribing calcitriol (activated vitamin D), which stimulates gut absorption of whatever calcium is taken.⁴⁹ If calcitriol is given, the patient must undergo cumbersome monitoring for hypercalcemia and hypercalciuria. Fortunately, it is unusual to require calcitriol unless the patient has significant structural gastrointestinal abnormalities such as gastric bypass or Crohn disease.

Although calcium and vitamin D are often recommended for prevention and treatment of osteoporosis, considerable controversy exists in terms of their safety and efficacy.¹ This article highlights the issues, referring  readers to reviews and meta-analyses for details and providing some practical advice for patients requiring supplementation.

CALCIUM INTAKE AND BONE DENSITY

Calcium enters the body through diet and supplementation. If intake is low, blood calcium levels fall, resulting in secondary hyperparathyroidism, which has 3 main effects:

• Increased fractional absorption of the calcium that is consumed
• Reduced urinary excretion of calcium
• Increased bone resorption, which releases calcium into the blood,² which explains the potential for the deleterious effect of deficient intake of calcium on bone.³

Based on the simple physiology outlined above, it seems logical that insufficient intake of calcium over time could lead to mobilization of calcium from bone, lower bone mineral density, and higher fracture risk.³ This topic has been reviewed by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis, and Musculoskeletal Diseases and the International Foundation for Osteoporosis.¹

Many lines of evidence suggest that low calcium intake adversely affects bone mineral density.¹ Low calcium intake has been associated with lower bone density in some cross-sectional studies,^4–6 though not all.⁷ Interventions to increase calcium intake in postmenopausal women have shown beneficial effects on bone density,^8–10 though in some studies the benefit was small and nonprogressive.¹¹ The question is whether this improvement in bone mineral density translates into fewer fractures.

Results from individual studies looking at fracture prevention through calcium supplementation have been conflicting,^10,12–14 and reviews and meta-analyses have summarized the data.^1,3,15 A recent review of these meta-analyses showed a small but significant reduction in some types of fracture.¹

Some speculate that the difficulty in demonstrating fracture efficacy might be due to imperfect compliance with calcium intake, and that the participants in the placebo groups often had fairly robust calcium intake from diet and off-study supplemental intake, which could reduce the sensitivity of studies to demonstrate the fracture benefit.^1,16

The US Preventive Services Task Force¹⁷ recommends that the general public not take supplemental calcium for skeletal health, but emphasizes that this recommendation does not apply to patients with osteoporosis. Most other official guidelines (eg, those of the Endocrine Society,¹⁸ American Association of Clinical Endocrinologists,¹⁹ Institute of Medicine,²⁰ and National Osteoporosis Foundation²¹) recommend adequate calcium intake to optimize skeletal health.

CALCIUM INTAKE AND CORONARY ARTERY DISEASE

Patients often wonder if the calcium in their supplements ends up in their coronary arteries rather than their bones. Although we once dismissed such concerns, several studies and meta-analyses have reported higher rates of cardiovascular disease with supplemental calcium use.^22–24 A proposed mechanism to explain this increased risk is that taking calcium supplements transiently raises the serum calcium level, resulting in calcium deposition in coronary arteries, accelerating atherosclerosis formation.²⁵

On the other hand, some studies and meta-analyses have not shown any increased risk of cardiovascular disease with calcium and vitamin D supplementation.^26,27 This subject has been reviewed by Harvey et al.¹

Our conclusions are as follows:

Patients should be told that the National Osteoporosis Foundation and the American Society for Preventive Cardiology released a statement in 2016 adopting the position that calcium intake from food and supplements should be considered safe from a cardiovascular perspective.²⁸

If patients want to avoid the possible increase in risk of cardiovascular disease due to calcium supplementation, they can optimize their calcium intake with dietary calcium. Observational studies that showed increased risk with supplemental calcium found no such increase in cardiovascular disease with a robust dietary intake of calcium.²⁹

This is not to say that patients should be encouraged to boost their dietary calcium intake and avoid heart disease by eating more cheese and ice cream, as these foods are high in saturated fats and cholesterol. Many dairy and nondairy sources of calcium do not contain these undesirable nutrients.

CALCIUM SUPPLEMENTATION AND NEPHROLITHIASIS

High dietary calcium intake has not been shown to increase the risk of kidney stones.

In the Nurses’ Health Study, the multivariate relative risk of stone formation was 0.65 (95% confidence interval [CI] 0.5–0.83) in those in the highest vs the lowest quintiles of dietary calcium intake.³⁰ In contrast, the relative risk of stones in those taking calcium supplements was 1.2 (CI 1.02–1.41),³⁰ although this higher risk was not seen in younger women (ages 27 to 44).³¹  

Similar results were seen in the Women’s Health Initiative, in which calcium carbonate and vitamin D supplements resulted in a relative increased risk of stone formation of 1.17 (95% CI 1.02–1.34) compared with women on placebo.¹²

Data from male stone-formers also suggests that high dietary calcium intake does not increase the risk of stones.³²

A theory to explain the difference between dietary and supplemental calcium with respect to stone formation is that dietary calcium binds to oxalate in the gut and reduces its absorption. The most common type of kidney stones are composed of calcium oxalate, and the oxalate, not the calcium, may be the real culprit. In contrast, calcium supplements are often taken between meals and therefore do not exert this protective effect and may be absorbed more rapidly and raise the serum calcium level more, which could lead to higher urinary calcium excretion.³³

CALCIUM INTAKE IN PATIENTS TAKING ANTIRESORPTIVE DRUGS

Patients often mistakenly think that calcium and vitamin D supplements are given for mild cases of bone loss, and that if their bone loss is significant enough to require a medication, then they no longer need calcium and vitamin D supplements. Most clinical trials showing bone mineral density and fracture benefit from antiresorptive therapy were in patients who were taking enough calcium and vitamin D, so the efficacy of antiresorptive therapy is most clear only when taking enough calcium and vitamin D.^34,35

Furthermore, patients with inadequate calcium and vitamin D intake essentially maintain their serum calcium levels by mobilizing calcium from bone; the combination of insufficient calcium intake and administration of agents that interfere with the ability to mobilize calcium from bone may put patients at risk of hypocalcemia.³⁶

OPTIMIZING INTAKE OF CALCIUM AND VITAMIN D

Diet is key to calcium intake. People should consume adequate amounts of calcium-rich foods regardless of whether they have a history of kidney stones, since robust dietary intake of calcium does not increase the risk of cardiovascular disease or kidney stones and may actually have a protective effect. We also remain skeptical of the concern that supplemental calcium increases the risk of cardiovascular disease.

We recommend a target total calcium intake from diet, and if necessary, supplements, of 1,000 to 1,200 mg daily, and not to worry about cardiovascular disease or kidney stones. A patient or clinician reluctant to push calcium intake that high with supplements might opt for a more conservative goal of 800 mg of calcium daily. This recommendation is based on data suggesting that in the presence of vitamin D sufficiency, calcium supplementation with 500 mg of calcium citrate does little for patients whose calcium intake is above 400 mg/day.³⁷

Vitamin D: How much do we need?

Regarding vitamin D intake, the Institute of Medicine recommends 600 to 800 IU to achieve a 25-hydroxyvitamin D level of 20 to 40 ng/mL.²⁰

The Endocrine Society recommends “at least” 600 to 800 IU, but says that 1,500 to 2,000 IU may be needed to get the 25-hydroxyvitamin D level to 30 to 60 ng/mL.¹⁸

The Institute of Medicine based its recommendation on randomized controlled trials that showed fewer fractures with vitamin D intakes of 600 to 800 IU/day.^13,14 Also, observational studies show little further reduction in fracture risk when the 25-hydroxyvitamin D levels rise above 20 ng/mL.³⁸ A case-control study found an association between 25-hydroxyvitamin D levels higher than 40 ng/mL and pancreatic cancer.³⁹

The Endocrine Society guidelines recommended higher intakes and levels of vitamin D because there are data suggesting that vitamin D levels higher than 30 ng/mL suppress parathyroid hormone levels further, which should favor less mobilization of bone.⁴⁰

Levels of 25-hydroxyvitamin D in people exposed to plenty of sunlight rarely go above 60 ng/mL, suggesting 60 ng/mL should be the upper limit of levels to target, and it is unlikely that such levels are harmful.⁴¹

Implementing either recommendation—a target 25-hydroxyvitamin D level of 20 to 40 ng/mL or 30 to 60 ng/mL—is reasonable.

CALCULATING A PATIENT’S DIETARY CALCIUM INTAKE

A detailed dietary history can be obtained by a dietitian, or by using the Calcium Calculator app supported by the International Osteoporosis Foundation.⁴³ However, dietary calcium intake can be assessed quickly. To approximate a patient’s total dietary calcium intake (in milligrams), we multiply the number of servings of dietary calcium by 300. A serving of dietary calcium is found in:

• 1 cup of milk, yogurt, calcium-fortified juice, almonds, cooked spinach, or collard greens
• 1.5 ounces of hard cheese
• 2 cups of ice cream, cottage cheese, or beans
• 4 ounces of tofu or canned fish with bones such as salmon or sardines.

Therefore, if a patient consumes 1 cup of milk daily and 1 cup of yogurt 3 times a week, she takes in an estimated 1.5 servings of dietary calcium daily, or 450 mg. What the patient does not receive in the diet should be made up with supplemental calcium.

CALCIUM SUPPLEMENTS

Calcium citrate has certain advantages as a supplement. Calcium carbonate requires gastric acidity to be absorbed and is therefore better absorbed if taken with meals; however, calcium citrate is equally well absorbed in the fasting or fed state and so can be taken without regard for achlorhydria or timing of meals.⁴⁴

Another potential advantage of calcium citrate is that it has never been shown to increase the risk of kidney stones the way calcium carbonate has.¹² Further, potassium citrate is a treatment for certain types of kidney stones,⁴⁵ and it is possible that when calcium is given as citrate there is less danger of kidney stones.⁴⁶ For these reasons, we generally recommend calcium citrate over other forms of calcium.

The brand of calcium citrate most readily available is Citracal, but any version of calcium citrate is acceptable.

SOURCES OF CONFUSION

Labels that describe calcium content of supplements are often misleading, and this lack of clarity can interfere with the patient’s ability to correctly identify how much calcium is in each pill.

Serving size. Whereas 1 serving of Caltrate is 1 pill, 1 serving of Tums or Citracal is 2 pills; for other brands a serving may be 3 or 4 pills.

Calcium salt vs elemental calcium. The amount of elemental calcium contained in different calcium salts varies according to the molecular weight of the salt: 1,000 mg of calcium carbonate has 400 mg of elemental calcium, while 1,000 mg of calcium citrate has 200 mg of elemental calcium.

When we recommend 1,000 to 1,200 mg of calcium daily, we mean the amount of elemental calcium. The label on calcium supplements usually indicates the amount of elemental calcium, but some have confusing information about the amount of calcium salt they contain. For instance, Tums lists the amount of calcium carbonate per pill on the top of the label, but elsewhere lists the amount of elemental calcium.

Same brand, different preparation. Some brands of calcium have more than 1 formulation, each with a different amount of calcium. For instance, Citracal has a maximum-strength 315-mg tablet and a “petite” 200-mg tablet. Careful reading of the label is required to make sure that the patient is getting the amount of calcium she thinks she is getting.

OPTIONS FOR THOSE WITH DIFFICULTY SWALLOWING LARGE PILLS

Many calcium pills are large and difficult to swallow. Patients often ask if calcium pills can be crushed, and the answer is that they certainly can, but this approach is cumbersome and usually results in patients eventually stopping calcium in frustration.

CALCIUM SUPPLEMENTS AND CONSTIPATION

Constipation is a common side effect of calcium supplementation.⁴⁷ Many patients report that they cannot take a calcium supplement because of constipation, or ask if there are calcium preparations that are less constipating than others.

There are ways of overcoming the constipating effects of calcium. Osmotic laxatives and stool softeners such as polyethylene glycol, magnesium citrate, and docusate sodium are safe and effective, although patients are often reluctant to take a medicine to combat the side effects from another medicine.

In such circumstances patients are often amenable to taking a combination product such as calcium with magnesium, since the cathartic effects of magnesium nicely counteract the constipating effects of calcium. This idea is exploited in antacids such as Rolaids, which are combinations of calcium carbonate and magnesium oxide that usually have no net effect on stool consistency.⁴⁷

Many patients believe that calcium must be combined with magnesium to be absorbed. Although there are no data to support this idea, a patient already harboring this misconception may be more amenable to calcium-magnesium combinations for the purpose of avoiding constipation.

If a patient cannot find a calcium preparation that she can take at the full recommended doses, we often suggest starting with a very small dose for 2 weeks, and then adjusting the dose upward every 2 weeks until reaching the maximum dose that the patient can tolerate. Even if the dose is well below recommended doses, most of the benefit of calcium is obtained by bringing total intake to more than 500 mg daily,³⁷ so continued use should be encouraged even when optimal targets cannot be sustained.

For patients who cannot tolerate enough calcium, we recommend being especially sure to optimize the vitamin D levels, since there are studies that suggest that secondary hyperparathyroidism mostly occurs in states of low calcium intake if vitamin D levels are insufficient.⁴⁸

If the patient has secondary hyperpara­thyroidism despite best attempts at supplementation with calcium and vitamin D, consider prescribing calcitriol (activated vitamin D), which stimulates gut absorption of whatever calcium is taken.⁴⁹ If calcitriol is given, the patient must undergo cumbersome monitoring for hypercalcemia and hypercalciuria. Fortunately, it is unusual to require calcitriol unless the patient has significant structural gastrointestinal abnormalities such as gastric bypass or Crohn disease.

Although calcium and vitamin D are often recommended for prevention and treatment of osteoporosis, considerable controversy exists in terms of their safety and efficacy.¹ This article highlights the issues, referring  readers to reviews and meta-analyses for details and providing some practical advice for patients requiring supplementation.

CALCIUM INTAKE AND BONE DENSITY

Calcium enters the body through diet and supplementation. If intake is low, blood calcium levels fall, resulting in secondary hyperparathyroidism, which has 3 main effects:

• Increased fractional absorption of the calcium that is consumed
• Reduced urinary excretion of calcium
• Increased bone resorption, which releases calcium into the blood,² which explains the potential for the deleterious effect of deficient intake of calcium on bone.³

Based on the simple physiology outlined above, it seems logical that insufficient intake of calcium over time could lead to mobilization of calcium from bone, lower bone mineral density, and higher fracture risk.³ This topic has been reviewed by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis, and Musculoskeletal Diseases and the International Foundation for Osteoporosis.¹

Many lines of evidence suggest that low calcium intake adversely affects bone mineral density.¹ Low calcium intake has been associated with lower bone density in some cross-sectional studies,^4–6 though not all.⁷ Interventions to increase calcium intake in postmenopausal women have shown beneficial effects on bone density,^8–10 though in some studies the benefit was small and nonprogressive.¹¹ The question is whether this improvement in bone mineral density translates into fewer fractures.

Results from individual studies looking at fracture prevention through calcium supplementation have been conflicting,^10,12–14 and reviews and meta-analyses have summarized the data.^1,3,15 A recent review of these meta-analyses showed a small but significant reduction in some types of fracture.¹

Some speculate that the difficulty in demonstrating fracture efficacy might be due to imperfect compliance with calcium intake, and that the participants in the placebo groups often had fairly robust calcium intake from diet and off-study supplemental intake, which could reduce the sensitivity of studies to demonstrate the fracture benefit.^1,16

The US Preventive Services Task Force¹⁷ recommends that the general public not take supplemental calcium for skeletal health, but emphasizes that this recommendation does not apply to patients with osteoporosis. Most other official guidelines (eg, those of the Endocrine Society,¹⁸ American Association of Clinical Endocrinologists,¹⁹ Institute of Medicine,²⁰ and National Osteoporosis Foundation²¹) recommend adequate calcium intake to optimize skeletal health.

CALCIUM INTAKE AND CORONARY ARTERY DISEASE

Patients often wonder if the calcium in their supplements ends up in their coronary arteries rather than their bones. Although we once dismissed such concerns, several studies and meta-analyses have reported higher rates of cardiovascular disease with supplemental calcium use.^22–24 A proposed mechanism to explain this increased risk is that taking calcium supplements transiently raises the serum calcium level, resulting in calcium deposition in coronary arteries, accelerating atherosclerosis formation.²⁵

On the other hand, some studies and meta-analyses have not shown any increased risk of cardiovascular disease with calcium and vitamin D supplementation.^26,27 This subject has been reviewed by Harvey et al.¹

Our conclusions are as follows:

Patients should be told that the National Osteoporosis Foundation and the American Society for Preventive Cardiology released a statement in 2016 adopting the position that calcium intake from food and supplements should be considered safe from a cardiovascular perspective.²⁸

If patients want to avoid the possible increase in risk of cardiovascular disease due to calcium supplementation, they can optimize their calcium intake with dietary calcium. Observational studies that showed increased risk with supplemental calcium found no such increase in cardiovascular disease with a robust dietary intake of calcium.²⁹

This is not to say that patients should be encouraged to boost their dietary calcium intake and avoid heart disease by eating more cheese and ice cream, as these foods are high in saturated fats and cholesterol. Many dairy and nondairy sources of calcium do not contain these undesirable nutrients.

CALCIUM SUPPLEMENTATION AND NEPHROLITHIASIS

High dietary calcium intake has not been shown to increase the risk of kidney stones.

In the Nurses’ Health Study, the multivariate relative risk of stone formation was 0.65 (95% confidence interval [CI] 0.5–0.83) in those in the highest vs the lowest quintiles of dietary calcium intake.³⁰ In contrast, the relative risk of stones in those taking calcium supplements was 1.2 (CI 1.02–1.41),³⁰ although this higher risk was not seen in younger women (ages 27 to 44).³¹  

Similar results were seen in the Women’s Health Initiative, in which calcium carbonate and vitamin D supplements resulted in a relative increased risk of stone formation of 1.17 (95% CI 1.02–1.34) compared with women on placebo.¹²

Data from male stone-formers also suggests that high dietary calcium intake does not increase the risk of stones.³²

A theory to explain the difference between dietary and supplemental calcium with respect to stone formation is that dietary calcium binds to oxalate in the gut and reduces its absorption. The most common type of kidney stones are composed of calcium oxalate, and the oxalate, not the calcium, may be the real culprit. In contrast, calcium supplements are often taken between meals and therefore do not exert this protective effect and may be absorbed more rapidly and raise the serum calcium level more, which could lead to higher urinary calcium excretion.³³

CALCIUM INTAKE IN PATIENTS TAKING ANTIRESORPTIVE DRUGS

Patients often mistakenly think that calcium and vitamin D supplements are given for mild cases of bone loss, and that if their bone loss is significant enough to require a medication, then they no longer need calcium and vitamin D supplements. Most clinical trials showing bone mineral density and fracture benefit from antiresorptive therapy were in patients who were taking enough calcium and vitamin D, so the efficacy of antiresorptive therapy is most clear only when taking enough calcium and vitamin D.^34,35

Furthermore, patients with inadequate calcium and vitamin D intake essentially maintain their serum calcium levels by mobilizing calcium from bone; the combination of insufficient calcium intake and administration of agents that interfere with the ability to mobilize calcium from bone may put patients at risk of hypocalcemia.³⁶

OPTIMIZING INTAKE OF CALCIUM AND VITAMIN D

Diet is key to calcium intake. People should consume adequate amounts of calcium-rich foods regardless of whether they have a history of kidney stones, since robust dietary intake of calcium does not increase the risk of cardiovascular disease or kidney stones and may actually have a protective effect. We also remain skeptical of the concern that supplemental calcium increases the risk of cardiovascular disease.

We recommend a target total calcium intake from diet, and if necessary, supplements, of 1,000 to 1,200 mg daily, and not to worry about cardiovascular disease or kidney stones. A patient or clinician reluctant to push calcium intake that high with supplements might opt for a more conservative goal of 800 mg of calcium daily. This recommendation is based on data suggesting that in the presence of vitamin D sufficiency, calcium supplementation with 500 mg of calcium citrate does little for patients whose calcium intake is above 400 mg/day.³⁷

Vitamin D: How much do we need?

Regarding vitamin D intake, the Institute of Medicine recommends 600 to 800 IU to achieve a 25-hydroxyvitamin D level of 20 to 40 ng/mL.²⁰

The Endocrine Society recommends “at least” 600 to 800 IU, but says that 1,500 to 2,000 IU may be needed to get the 25-hydroxyvitamin D level to 30 to 60 ng/mL.¹⁸

The Institute of Medicine based its recommendation on randomized controlled trials that showed fewer fractures with vitamin D intakes of 600 to 800 IU/day.^13,14 Also, observational studies show little further reduction in fracture risk when the 25-hydroxyvitamin D levels rise above 20 ng/mL.³⁸ A case-control study found an association between 25-hydroxyvitamin D levels higher than 40 ng/mL and pancreatic cancer.³⁹

The Endocrine Society guidelines recommended higher intakes and levels of vitamin D because there are data suggesting that vitamin D levels higher than 30 ng/mL suppress parathyroid hormone levels further, which should favor less mobilization of bone.⁴⁰

Levels of 25-hydroxyvitamin D in people exposed to plenty of sunlight rarely go above 60 ng/mL, suggesting 60 ng/mL should be the upper limit of levels to target, and it is unlikely that such levels are harmful.⁴¹

Implementing either recommendation—a target 25-hydroxyvitamin D level of 20 to 40 ng/mL or 30 to 60 ng/mL—is reasonable.

CALCULATING A PATIENT’S DIETARY CALCIUM INTAKE

A detailed dietary history can be obtained by a dietitian, or by using the Calcium Calculator app supported by the International Osteoporosis Foundation.⁴³ However, dietary calcium intake can be assessed quickly. To approximate a patient’s total dietary calcium intake (in milligrams), we multiply the number of servings of dietary calcium by 300. A serving of dietary calcium is found in:

• 1 cup of milk, yogurt, calcium-fortified juice, almonds, cooked spinach, or collard greens
• 1.5 ounces of hard cheese
• 2 cups of ice cream, cottage cheese, or beans
• 4 ounces of tofu or canned fish with bones such as salmon or sardines.

Therefore, if a patient consumes 1 cup of milk daily and 1 cup of yogurt 3 times a week, she takes in an estimated 1.5 servings of dietary calcium daily, or 450 mg. What the patient does not receive in the diet should be made up with supplemental calcium.

CALCIUM SUPPLEMENTS

Calcium citrate has certain advantages as a supplement. Calcium carbonate requires gastric acidity to be absorbed and is therefore better absorbed if taken with meals; however, calcium citrate is equally well absorbed in the fasting or fed state and so can be taken without regard for achlorhydria or timing of meals.⁴⁴

Another potential advantage of calcium citrate is that it has never been shown to increase the risk of kidney stones the way calcium carbonate has.¹² Further, potassium citrate is a treatment for certain types of kidney stones,⁴⁵ and it is possible that when calcium is given as citrate there is less danger of kidney stones.⁴⁶ For these reasons, we generally recommend calcium citrate over other forms of calcium.

The brand of calcium citrate most readily available is Citracal, but any version of calcium citrate is acceptable.

SOURCES OF CONFUSION

Labels that describe calcium content of supplements are often misleading, and this lack of clarity can interfere with the patient’s ability to correctly identify how much calcium is in each pill.

Serving size. Whereas 1 serving of Caltrate is 1 pill, 1 serving of Tums or Citracal is 2 pills; for other brands a serving may be 3 or 4 pills.

Calcium salt vs elemental calcium. The amount of elemental calcium contained in different calcium salts varies according to the molecular weight of the salt: 1,000 mg of calcium carbonate has 400 mg of elemental calcium, while 1,000 mg of calcium citrate has 200 mg of elemental calcium.

When we recommend 1,000 to 1,200 mg of calcium daily, we mean the amount of elemental calcium. The label on calcium supplements usually indicates the amount of elemental calcium, but some have confusing information about the amount of calcium salt they contain. For instance, Tums lists the amount of calcium carbonate per pill on the top of the label, but elsewhere lists the amount of elemental calcium.

Same brand, different preparation. Some brands of calcium have more than 1 formulation, each with a different amount of calcium. For instance, Citracal has a maximum-strength 315-mg tablet and a “petite” 200-mg tablet. Careful reading of the label is required to make sure that the patient is getting the amount of calcium she thinks she is getting.

OPTIONS FOR THOSE WITH DIFFICULTY SWALLOWING LARGE PILLS

Many calcium pills are large and difficult to swallow. Patients often ask if calcium pills can be crushed, and the answer is that they certainly can, but this approach is cumbersome and usually results in patients eventually stopping calcium in frustration.

CALCIUM SUPPLEMENTS AND CONSTIPATION

Constipation is a common side effect of calcium supplementation.⁴⁷ Many patients report that they cannot take a calcium supplement because of constipation, or ask if there are calcium preparations that are less constipating than others.

There are ways of overcoming the constipating effects of calcium. Osmotic laxatives and stool softeners such as polyethylene glycol, magnesium citrate, and docusate sodium are safe and effective, although patients are often reluctant to take a medicine to combat the side effects from another medicine.

In such circumstances patients are often amenable to taking a combination product such as calcium with magnesium, since the cathartic effects of magnesium nicely counteract the constipating effects of calcium. This idea is exploited in antacids such as Rolaids, which are combinations of calcium carbonate and magnesium oxide that usually have no net effect on stool consistency.⁴⁷

Many patients believe that calcium must be combined with magnesium to be absorbed. Although there are no data to support this idea, a patient already harboring this misconception may be more amenable to calcium-magnesium combinations for the purpose of avoiding constipation.

If a patient cannot find a calcium preparation that she can take at the full recommended doses, we often suggest starting with a very small dose for 2 weeks, and then adjusting the dose upward every 2 weeks until reaching the maximum dose that the patient can tolerate. Even if the dose is well below recommended doses, most of the benefit of calcium is obtained by bringing total intake to more than 500 mg daily,³⁷ so continued use should be encouraged even when optimal targets cannot be sustained.

For patients who cannot tolerate enough calcium, we recommend being especially sure to optimize the vitamin D levels, since there are studies that suggest that secondary hyperparathyroidism mostly occurs in states of low calcium intake if vitamin D levels are insufficient.⁴⁸

If the patient has secondary hyperpara­thyroidism despite best attempts at supplementation with calcium and vitamin D, consider prescribing calcitriol (activated vitamin D), which stimulates gut absorption of whatever calcium is taken.⁴⁹ If calcitriol is given, the patient must undergo cumbersome monitoring for hypercalcemia and hypercalciuria. Fortunately, it is unusual to require calcitriol unless the patient has significant structural gastrointestinal abnormalities such as gastric bypass or Crohn disease.

With so much emphasis on clinical trials, evidence-based joint decision-making, and comparative-benefit studies when choosing treatment, the growth of the supplement market is a strong comment on the perceived and often real failings of traditional therapies. It also reflects our apparent failure as a profession to educate ourselves and the public about the difference between anecdote-based belief and clinical trial-based confidence, the difference between evidence and innuendo, and, equally important, the limitations of applying population-based clinical trial data to an individual patient.

Which brings me to the discussion of calcium and vitamin D supplementation by Drs. Kilim and Rosen in this issue of the Journal. We know a lot about calcium homeostasis and the role vitamin D plays in regulating circulating calcium levels. Only a small fraction of the calcium in the body circulates (most is in our skeleton), and likely only about 1% is truly exchangeable. But the circulating free calcium level is tightly controlled, as the function of our neuromuscular system, brain, and heart depend on keeping intra- and extracellular calcium levels within precise limits. If necessary, our bodies maintain stable levels of circulating free calcium at the expense of leaching calcium from our bones, placing us at risk of potentially fatal fractures. Thus was born the concept of guaranteeing adequate calcium stores through calcium supplementation.

Control of the free calcium level is not simple. There are several interrelated sensing and modulatory pathways, eg:

• Gut absorption, which is affected by the total gut load of calcium, intestinal integrity, and the specific ingested foodstuffs, and likely by our microbiome
• The parathyroid hormone (PTH) level, which directly or indirectly affects calcium absorption, the calcium-phosphate ratio, and thus, extraskeletal calcium localization and bone calcium content
• Vitamin D, with its many effects after interorgan multistep activation.

Despite this knowledge of calcium metabolism and the intricate cross-talk between the different pathways, I do not believe we truly understand how to determine the amount of dietary and supplemental calcium or vitamin D that is ideal for a given patient. I also do not believe we know with certainty what is the “normal” or ideal 25-hydroxyvitamin D level in that same patient: note the different ranges of normal proposed by different expert working groups.

How do we know when there is insufficient calcium in our diet? The total or free circulating calcium level is far too insensitive and, as noted above, complex homeostatic mechanisms are always trying to maintain an appropriate physiologic calcium level, whatever the intake. Urinary calcium excretion does not necessarily equal the ingested calcium load; there are too many factors influencing renal calcium excretion. Gastrointestinal excretion is also variable and complex. We know when the vitamin D level is functionally much too low, as the PTH level begins to rise; but the “normal” PTH range is wide, and the slope of the relationship between vitamin D and PTH is affected by many factors.

Additionally, accumulating information suggests that vitamin D metabolites significantly affect immune regulation and the onset and expression of a number of organ-specific and systemic autoimmune disorders. Further, the ideal 25-hydroxyvitamin D level for a healthy immune system is not known. Does the body have to compromise something in reconciling the target for a healthy skeleton and the target for a healthy immune system, which may conceivably be different? But importantly, this new knowledge does not imply that vitamin D supplementation will reduce the pain and symptoms from inflammatory arthritis or noninflammatory fibromyalgia.

Despite these many areas of uncertainty, Kilim and Rosen focus on bone health, summarize the wealth of accumulated data, and provide practical management advice we can use in the clinic. But if we struggle so much to know the correct way to manage calcium and vitamin D supplementation in our patients, it is little surprise that most of us struggle even more when asked about using supplements for which the biology is far less understood. As a medical community, we need to uniformly address this lack of understanding wherever it exists. Believing in a supplement or a treatment is not the same as understanding it or having strong evidence about its efficacy and safety.

With so much emphasis on clinical trials, evidence-based joint decision-making, and comparative-benefit studies when choosing treatment, the growth of the supplement market is a strong comment on the perceived and often real failings of traditional therapies. It also reflects our apparent failure as a profession to educate ourselves and the public about the difference between anecdote-based belief and clinical trial-based confidence, the difference between evidence and innuendo, and, equally important, the limitations of applying population-based clinical trial data to an individual patient.

Which brings me to the discussion of calcium and vitamin D supplementation by Drs. Kilim and Rosen in this issue of the Journal. We know a lot about calcium homeostasis and the role vitamin D plays in regulating circulating calcium levels. Only a small fraction of the calcium in the body circulates (most is in our skeleton), and likely only about 1% is truly exchangeable. But the circulating free calcium level is tightly controlled, as the function of our neuromuscular system, brain, and heart depend on keeping intra- and extracellular calcium levels within precise limits. If necessary, our bodies maintain stable levels of circulating free calcium at the expense of leaching calcium from our bones, placing us at risk of potentially fatal fractures. Thus was born the concept of guaranteeing adequate calcium stores through calcium supplementation.

Control of the free calcium level is not simple. There are several interrelated sensing and modulatory pathways, eg:

• Gut absorption, which is affected by the total gut load of calcium, intestinal integrity, and the specific ingested foodstuffs, and likely by our microbiome
• The parathyroid hormone (PTH) level, which directly or indirectly affects calcium absorption, the calcium-phosphate ratio, and thus, extraskeletal calcium localization and bone calcium content
• Vitamin D, with its many effects after interorgan multistep activation.

Despite this knowledge of calcium metabolism and the intricate cross-talk between the different pathways, I do not believe we truly understand how to determine the amount of dietary and supplemental calcium or vitamin D that is ideal for a given patient. I also do not believe we know with certainty what is the “normal” or ideal 25-hydroxyvitamin D level in that same patient: note the different ranges of normal proposed by different expert working groups.

How do we know when there is insufficient calcium in our diet? The total or free circulating calcium level is far too insensitive and, as noted above, complex homeostatic mechanisms are always trying to maintain an appropriate physiologic calcium level, whatever the intake. Urinary calcium excretion does not necessarily equal the ingested calcium load; there are too many factors influencing renal calcium excretion. Gastrointestinal excretion is also variable and complex. We know when the vitamin D level is functionally much too low, as the PTH level begins to rise; but the “normal” PTH range is wide, and the slope of the relationship between vitamin D and PTH is affected by many factors.

Additionally, accumulating information suggests that vitamin D metabolites significantly affect immune regulation and the onset and expression of a number of organ-specific and systemic autoimmune disorders. Further, the ideal 25-hydroxyvitamin D level for a healthy immune system is not known. Does the body have to compromise something in reconciling the target for a healthy skeleton and the target for a healthy immune system, which may conceivably be different? But importantly, this new knowledge does not imply that vitamin D supplementation will reduce the pain and symptoms from inflammatory arthritis or noninflammatory fibromyalgia.

Despite these many areas of uncertainty, Kilim and Rosen focus on bone health, summarize the wealth of accumulated data, and provide practical management advice we can use in the clinic. But if we struggle so much to know the correct way to manage calcium and vitamin D supplementation in our patients, it is little surprise that most of us struggle even more when asked about using supplements for which the biology is far less understood. As a medical community, we need to uniformly address this lack of understanding wherever it exists. Believing in a supplement or a treatment is not the same as understanding it or having strong evidence about its efficacy and safety.

With so much emphasis on clinical trials, evidence-based joint decision-making, and comparative-benefit studies when choosing treatment, the growth of the supplement market is a strong comment on the perceived and often real failings of traditional therapies. It also reflects our apparent failure as a profession to educate ourselves and the public about the difference between anecdote-based belief and clinical trial-based confidence, the difference between evidence and innuendo, and, equally important, the limitations of applying population-based clinical trial data to an individual patient.

Which brings me to the discussion of calcium and vitamin D supplementation by Drs. Kilim and Rosen in this issue of the Journal. We know a lot about calcium homeostasis and the role vitamin D plays in regulating circulating calcium levels. Only a small fraction of the calcium in the body circulates (most is in our skeleton), and likely only about 1% is truly exchangeable. But the circulating free calcium level is tightly controlled, as the function of our neuromuscular system, brain, and heart depend on keeping intra- and extracellular calcium levels within precise limits. If necessary, our bodies maintain stable levels of circulating free calcium at the expense of leaching calcium from our bones, placing us at risk of potentially fatal fractures. Thus was born the concept of guaranteeing adequate calcium stores through calcium supplementation.

Control of the free calcium level is not simple. There are several interrelated sensing and modulatory pathways, eg:

• Gut absorption, which is affected by the total gut load of calcium, intestinal integrity, and the specific ingested foodstuffs, and likely by our microbiome
• The parathyroid hormone (PTH) level, which directly or indirectly affects calcium absorption, the calcium-phosphate ratio, and thus, extraskeletal calcium localization and bone calcium content
• Vitamin D, with its many effects after interorgan multistep activation.

Despite this knowledge of calcium metabolism and the intricate cross-talk between the different pathways, I do not believe we truly understand how to determine the amount of dietary and supplemental calcium or vitamin D that is ideal for a given patient. I also do not believe we know with certainty what is the “normal” or ideal 25-hydroxyvitamin D level in that same patient: note the different ranges of normal proposed by different expert working groups.

How do we know when there is insufficient calcium in our diet? The total or free circulating calcium level is far too insensitive and, as noted above, complex homeostatic mechanisms are always trying to maintain an appropriate physiologic calcium level, whatever the intake. Urinary calcium excretion does not necessarily equal the ingested calcium load; there are too many factors influencing renal calcium excretion. Gastrointestinal excretion is also variable and complex. We know when the vitamin D level is functionally much too low, as the PTH level begins to rise; but the “normal” PTH range is wide, and the slope of the relationship between vitamin D and PTH is affected by many factors.

Additionally, accumulating information suggests that vitamin D metabolites significantly affect immune regulation and the onset and expression of a number of organ-specific and systemic autoimmune disorders. Further, the ideal 25-hydroxyvitamin D level for a healthy immune system is not known. Does the body have to compromise something in reconciling the target for a healthy skeleton and the target for a healthy immune system, which may conceivably be different? But importantly, this new knowledge does not imply that vitamin D supplementation will reduce the pain and symptoms from inflammatory arthritis or noninflammatory fibromyalgia.

Despite these many areas of uncertainty, Kilim and Rosen focus on bone health, summarize the wealth of accumulated data, and provide practical management advice we can use in the clinic. But if we struggle so much to know the correct way to manage calcium and vitamin D supplementation in our patients, it is little surprise that most of us struggle even more when asked about using supplements for which the biology is far less understood. As a medical community, we need to uniformly address this lack of understanding wherever it exists. Believing in a supplement or a treatment is not the same as understanding it or having strong evidence about its efficacy and safety.

Cardiac rehabilitation has a class 1 indication (ie, strong recommendation) after heart surgery, myocardial infarction, or coronary intervention, and for stable angina or peripheral artery disease. It has a class 2a indication (ie, moderate recommendation) for stable systolic heart failure. Yet it is still under­utilized despite its demonstrated benefits, endorsement by most recognized cardiovascular societies, and coverage by the US Centers for Medicare and Medicaid Services (CMS).

Here, we review cardiac rehabilitation—its benefits, appropriate indications, barriers to referral and enrollment, and efforts to increase its use.

EXERCISE: SLOW TO BE ADOPTED

In 1772, William Heberden (also remembered today for describing swelling of the distal interphalangeal joints in osteoarthritis) described¹ a patient with angina pectoris who “set himself a task of sawing wood for half an hour every day, and was nearly cured.”

Despite early clues, it would be some time before the medical community would recognize the benefits of exercise for cardiovascular health. Before the 1930s, immobilization and extended bedrest were encouraged for up to 6 weeks after a cardiovascular event, leading to significant deconditioning.² Things slowly began to change in the 1940s with Levine’s introduction of up-to-chair therapy,³ and short daily walks were introduced in the 1950s. Over time, the link between a sedentary lifestyle and cardiovascular disease was studied and led to greater investigation into the benefits of exercise, propelling us into the modern era.^4,5

CARDIAC REHABILITATION: COMPREHENSIVE RISK REDUCTION

The American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) defines cardiac rehabilitation as the provision of comprehensive long-term services involving medical evaluation, prescriptive exercise, cardiac risk-factor modification, education, counseling, and behavioral interventions.⁶ CMS defines it as a physician-supervised program that furnishes physician-prescribed exercise, cardiac risk-factor modification (including education, counseling, and behavioral intervention), psychosocial assessment, outcomes assessment, and other items and services.⁷

In general, most cardiac rehabilitation programs provide medically supervised exercise and patient education designed to improve cardiac health and functional status. Risk factors are targeted to reduce disability and rates of morbidity and mortality, to improve functional capacity, and to alleviate activity-related symptoms.

FROM HOSPITAL TO SELF-MAINTENANCE

Phase 1: Inpatient rehabilitation

Phase 1 typically takes place in the inpatient setting, often after open heart surgery (eg, coronary artery bypass grafting, valve repair or replacement, heart transplant), myocardial infarction, or percutaneous coronary intervention. This phase may last only a few days, especially in the current era of short hospital stays.

During phase 1, patients discuss their health situation and goals with their primary provider or cardiologist and receive education about recovery and cardiovascular risk factors. Early mobilization to prepare for discharge and to resume simple activities of daily living is emphasized. Depending on the institution, phase 1 exercise may involve simple ambulation on the ward or using equipment such as a stationary bike or treadmill.⁶ Phase 2 enrollment ideally is set up before discharge.

Phase 2: Limited-time outpatient rehabilitation

Phase 2 traditionally takes place in a hospital-based outpatient facility and consists of a physician-supervised multidisciplinary program. Growing evidence shows that home-based cardiac rehabilitation may be as effective as a medical facility-based program and should be an option for patients who have difficulty getting access to a traditional program.⁸

A phase 2 program takes a threefold approach, consisting of exercise, aggressive risk-factor modification, and education classes. A Cochrane review⁹ included programs that also incorporated behavioral modification and psychosocial support as a means of secondary prevention, underscoring the evolving definition of cardiac rehabilitation.

During the initial phase 2 visit, an individualized treatment plan is developed, incorporating an exercise prescription and realistic goals for secondary prevention. Sessions typically take place 3 times a week for up to 36 sessions; usually, options are available for less frequent weekly attendance for a longer period to achieve a full course. In some cases, patients may qualify for up to 72 sessions, particularly if they have not progressed as expected.

Exercise. As part of the initial evaluation, AACVPR guidelines⁶ suggest an exercise test­—eg, a symptom-limited exercise stress test, a 6-minute walk test, or use of a Rating of Perceived Exertion scale. Prescribed exercise generally targets moderate activity in the range of 50% to 70% of peak estimated functional capacity. In the appropriate clinical context, high-functioning patients can be offered high-intensity interval training instead of moderate exercise, as they confer similar benefits.¹⁰

Risk-factor reduction. Comprehensive risk-factor reduction can address smoking, hypertension, high cholesterol, diabetes, obesity, and diet, as well as psychosocial issues such as stress, anxiety, depression, and alcohol use. Sexual activity counseling may also be included.

Education classes are aimed at helping patients understand cardiovascular disease and empowering them to manage their medical treatment and lifestyle modifications.⁶

Phase 3: Lifetime maintenance

In phase 3, patients independently continue risk-factor modification and physical activity without cardiac monitoring. Most cardiac rehabilitation programs offer transition-to-maintenance classes after completion of phase 2; this may be a welcome option, particularly for those who have developed a good routine and rapport with the staff and other participants. Others may opt for an independent program, using their own home equipment or a local health club.

EXERCISE: MOSTLY SAFE, WITH PROVEN BENEFITS

The safety of cardiac rehabilitation is well established, with a low risk of major cardiovascular complications. A US study in the early 1980s of 167 cardiac rehabilitation programs found 1 cardiac arrest for every 111,996 exercise hours, 1 myocardial infarction per 293,990 exercise hours, and 1 fatality per 783,972 exercise hours.¹¹ A 2006 study of more than 65 cardiac rehabilitation centers in France found 1 cardiac event per 8,484 exercise tests and 1.3 cardiac arrests per 1 million exercise hours.¹²

The benefits of cardiac rehabilitation are numerous and substantial.^9,13–17 A 2016 Cochrane review and meta-analysis of 63 randomized controlled trials with 14,486 participants found a reduced rate of cardiovascular mortality (relative risk [RR] 0.74, 95% confidence interval [CI] 0.64–0.86), with a number needed to treat of 37, and fewer hospital re­admissions (RR 0.82, 95% CI 0.70–0.96).⁹

Reductions in mortality rates are dose-dependent. A study of more than 30,000 Medicare beneficiaries who participated in cardiac rehabilitation found that those who attended more sessions had a lower rate of morbidity and death at 4 years, particularly if they participated in more than 11 sessions. Those who attended the full 36 sessions had a mortality rate 47% lower than those who attended a single session.¹⁷ There was a 15% reduction in mortality for those who attended 36 sessions compared with 24 sessions, a 28% lower risk with attending 36 sessions compared with 12. After adjustment, each additional 6 sessions was associated with a 6% reduction in mortality. The curves continued to separate up to 4 years.

The benefits of cardiac rehabilitation go beyond risk reduction and include improved functional capacity, greater ease with activities of daily living, and improved quality of life.⁹ Patients receive structure and support from the management team and other participants, which may provide an additional layer of friendship and psychosocial support for making lifestyle changes.

Is the overall mortality rate improved?

In the modern era, with access to optimal medical therapy and drug-eluting stents, one might expect only small additional benefit from cardiac rehabilitation. The 2016 Cochrane review and meta-analysis found that although cardiac rehabilitation contributed to improved cardiovascular mortality rates and health-related quality of life, no significant reduction was detected in the rate of death from all causes.⁸ But the analysis did not necessarily support removing the claim of reduced all-cause mortality for cardiac rehabilitation: only randomized controlled trials were examined, and the quality of evidence for each outcome was deemed to be low to moderate because of a general paucity of reports, including many small trials that followed patients for less than 12 months.

A large cohort analysis¹⁵ with more than 73,000 patients who had undergone cardiac rehabilitation found a relative reduction in mortality rate of 58% at 1 year and 21% to 34% at 5 years, with elderly women gaining the most benefit. In the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial, with more than 2,300 patients followed for a median of 2.5 years, exercise training for heart failure was associated with reduced rates of all-cause mortality or hospitalization (HR 0.89, 95% CI 0.81–0.99; P = .03) and of cardiovascular mortality or heart failure hospitalization (HR 0.85, 95% CI 0.74–0.99; P = .03).¹⁸

Regardless of the precise reduction in all-cause mortality, the cardiovascular and health-quality outcomes of cardiac rehabilitation clearly indicate benefit. More trials with follow-up longer than 1 year are needed to definitively determine the impact of cardiac rehabilitation on the all-cause mortality rate.

WHO SHOULD BE OFFERED CARDIAC REHABILITATION?

The 2006 CMS coverage criteria listed the indications for cardiac rehabilitation as myocardial infarction within the preceding 12 months, coronary artery bypass surgery, stable angina pectoris, heart valve repair or replacement, percutaneous coronary intervention, and heart or heart-lung transplant.

In 2014, stable chronic systolic heart failure was added to the list (Table 2). Qualifications include New York Heart Association class II (mild symptoms, slight limitation of activity) to class IV (severe limitations, symptoms at rest), an ejection fraction of 35% or less, and being stable on optimal medical therapy for at least 6 weeks.

In 2017, CMS approved supervised exercise therapy for peripheral arterial disease. Supervised exercise has a class 1 recommendation by the American Heart Association and American College of Cardiology for treating intermittent claudication. Supervised exercise therapy can increase walking distance by 180% and is superior to medical therapy alone. Unsupervised exercise has a class 2b recommendation.^19,20

Other patients may not qualify for phase 2 cardiac rehabilitation according to CMS or private insurance but could benefit from an exercise prescription and enrollment in a local phase 3 or home exercise program. Indications might include diabetes, obesity, metabolic syndrome, atrial fibrillation, postural orthostatic tachycardia syndrome, and nonalcoholic steatohepatitis. The benefits of cardiac rehabilitation after newer, less-invasive procedures for transcatheter valve repair and replacement are not well established, and more research is needed in this area.

WHEN TO REFER

Ades et al have defined cardiac rehabilitation referral as a combination of electronic medical records order, patient-physician discussion, and receipt of an order by a cardiac rehabilitation program.²¹

Ideally, referral for outpatient cardiac rehabilitation should take place at the time of hospital discharge. The AACVPR endorses a “cardiovascular continuum of care” model that emphasizes a smooth transition from inpatient to outpatient programs.⁶ Inpatient referral is a strong predictor of cardiac rehabilitation enrollment, and lack of referral in phase 1 negatively affects enrollment rates.

Depending on the diagnosis, US and Canadian guidelines recommend cardiac rehabilitation starting within 1 to 4 weeks of the index event, with acceptable wait times up to 60 days.^6,22 In the United Kingdom, referral is recommended within 24 hours of patient eligibility; assessment for a cardiovascular prevention and rehabilitation program, with a defined pathway and individual goals, is expected to be completed within 10 working days of referral.²³ Such a standard is difficult to meet in the United States, where the time from hospital discharge to cardiac rehabilitation program enrollment averages 35 days.^24,25

After an uncomplicated myocardial infarction or percutaneous coronary intervention, patients with a normal or mildly reduced left ventricular ejection fraction should start outpatient cardiac rehabilitation within 14 days of the index event. For such cases, cardiac rehabilitation has been shown to be safe within 1 to 2 weeks of hospital discharge and is associated with increased participation rates.

REHABILITATION IS STILL UNDERUSED

Despite its significant benefits, cardiac rehabilitation is underused for many reasons.

Referral rates vary

A study using the 1997 Medicare claims data­base showed national referral rates of only 14% after myocardial infarction and 31% after coronary artery bypass grafting.³¹

A later study using the National Cardiovascular Data Registry between 2009 and 2017 found that the situation had improved, with a referral rate of about 60% for patients undergoing percutaneous coronary intervention.³² Nevertheless, referral rates for cardiac rehabilitation remain highly variable and still lag behind other CMS quality measures for optimal medical therapy after acute myocardial infarction (Figure 1). Factors associated with higher referral rates included ST-segment elevation myocardial infarction, non-ST-segment elevation myocardial infarction, care in a high-volume center for percutaneous coronary intervention, and care in a private or community hospital in a Midwestern state. Small Midwestern hospitals generally had referral rates of over 80%, while major teaching hospitals and hospital systems on the East Coast and the West Coast had referral rates of less than 20%. Unlike some studies, this study found that insurance status had little bearing on referral rates.

Other studies found lower referral rates for women and patients with comorbidities such as previous coronary artery bypass grafting, diabetes, and heart failure.^33,34

In the United Kingdom, patients with heart failure made up only 5% of patients in cardiac rehabilitation; only 7% to 20% of patients with a heart failure diagnosis were referred to cardiac rehabilitation from general and cardiology wards.³⁵

Enrollment, completion rates even lower

Rates of referral for cardiac rehabilitation do not equate to rates of enrollment or participation. Enrollment was 50% in the United Kingdom in 2016.³⁵ A 2015 US study evaluated 58,269 older patients eligible for cardiac rehabilitation after acute myocardial infarction;  62% were referred for cardiac rehabilitation at the time of discharge, but only 23% of the total attended at least 1 session, and just 5% of the total completed 36 or more sessions.³⁶

BARRIERS, OPPORTUNITIES TO IMPROVE

The underuse of cardiac rehabilitation in the United States has led to an American Heart Association presidential advisory on the referral, enrollment, and delivery of cardiac rehabilitation.³⁴ Dozens of barriers are mentioned, with several standing out as having the largest impact: lack of physician referral, weak endorsement by the prescribing provider, female sex of patients, lack of program availability, work-related hardship, low socioeconomic status, and lack of or limited healthcare insurance. Copayments have also become a major barrier, often ranging from $20 to $40 per session for patients with Medicare.

The Million Hearts Initiative has established a goal of 70% cardiac rehabilitation compliance for eligible patients by 2022, a goal they estimate could save 25,000 lives and prevent 180,000 hospitalizations annually.²¹

Lack of physician awareness and lack of referral may be the most modifiable factors with the capacity to have the largest impact. Increasing physician awareness is a top priority not only for primary care providers, but also for cardiologists. In 2014, CMS made referral for cardiac rehabilitation a quality measure that is trackable and reportable. CMS has also proposed models that would incentivize participation by increasing reimbursement for services provided, but these models have been halted.

Additional efforts to increase cardiac rehabilitation referral and participation include automated order sets, increased caregiver education, and early morning or late evening classes, single-sex classes, home or mobile-based exercise programs, and parking and transportation assistance.³⁴ Grace et al³⁷ reported that referral rates rose to 86% when a cardiac rehabilitation order was integrated into the electronic medical record and combined with a hospital liaison to educate patients about their need for cardiac rehabilitation. Lowering patient copayments would also be a good idea. We have recently seen some creative ways to reduce copayments, including philanthropy and grants.

Cardiac rehabilitation has a class 1 indication (ie, strong recommendation) after heart surgery, myocardial infarction, or coronary intervention, and for stable angina or peripheral artery disease. It has a class 2a indication (ie, moderate recommendation) for stable systolic heart failure. Yet it is still under­utilized despite its demonstrated benefits, endorsement by most recognized cardiovascular societies, and coverage by the US Centers for Medicare and Medicaid Services (CMS).

Here, we review cardiac rehabilitation—its benefits, appropriate indications, barriers to referral and enrollment, and efforts to increase its use.

EXERCISE: SLOW TO BE ADOPTED

In 1772, William Heberden (also remembered today for describing swelling of the distal interphalangeal joints in osteoarthritis) described¹ a patient with angina pectoris who “set himself a task of sawing wood for half an hour every day, and was nearly cured.”

Despite early clues, it would be some time before the medical community would recognize the benefits of exercise for cardiovascular health. Before the 1930s, immobilization and extended bedrest were encouraged for up to 6 weeks after a cardiovascular event, leading to significant deconditioning.² Things slowly began to change in the 1940s with Levine’s introduction of up-to-chair therapy,³ and short daily walks were introduced in the 1950s. Over time, the link between a sedentary lifestyle and cardiovascular disease was studied and led to greater investigation into the benefits of exercise, propelling us into the modern era.^4,5

CARDIAC REHABILITATION: COMPREHENSIVE RISK REDUCTION

The American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) defines cardiac rehabilitation as the provision of comprehensive long-term services involving medical evaluation, prescriptive exercise, cardiac risk-factor modification, education, counseling, and behavioral interventions.⁶ CMS defines it as a physician-supervised program that furnishes physician-prescribed exercise, cardiac risk-factor modification (including education, counseling, and behavioral intervention), psychosocial assessment, outcomes assessment, and other items and services.⁷

In general, most cardiac rehabilitation programs provide medically supervised exercise and patient education designed to improve cardiac health and functional status. Risk factors are targeted to reduce disability and rates of morbidity and mortality, to improve functional capacity, and to alleviate activity-related symptoms.

FROM HOSPITAL TO SELF-MAINTENANCE

Phase 1: Inpatient rehabilitation

Phase 1 typically takes place in the inpatient setting, often after open heart surgery (eg, coronary artery bypass grafting, valve repair or replacement, heart transplant), myocardial infarction, or percutaneous coronary intervention. This phase may last only a few days, especially in the current era of short hospital stays.

During phase 1, patients discuss their health situation and goals with their primary provider or cardiologist and receive education about recovery and cardiovascular risk factors. Early mobilization to prepare for discharge and to resume simple activities of daily living is emphasized. Depending on the institution, phase 1 exercise may involve simple ambulation on the ward or using equipment such as a stationary bike or treadmill.⁶ Phase 2 enrollment ideally is set up before discharge.

Phase 2: Limited-time outpatient rehabilitation

Phase 2 traditionally takes place in a hospital-based outpatient facility and consists of a physician-supervised multidisciplinary program. Growing evidence shows that home-based cardiac rehabilitation may be as effective as a medical facility-based program and should be an option for patients who have difficulty getting access to a traditional program.⁸

A phase 2 program takes a threefold approach, consisting of exercise, aggressive risk-factor modification, and education classes. A Cochrane review⁹ included programs that also incorporated behavioral modification and psychosocial support as a means of secondary prevention, underscoring the evolving definition of cardiac rehabilitation.

During the initial phase 2 visit, an individualized treatment plan is developed, incorporating an exercise prescription and realistic goals for secondary prevention. Sessions typically take place 3 times a week for up to 36 sessions; usually, options are available for less frequent weekly attendance for a longer period to achieve a full course. In some cases, patients may qualify for up to 72 sessions, particularly if they have not progressed as expected.

Exercise. As part of the initial evaluation, AACVPR guidelines⁶ suggest an exercise test­—eg, a symptom-limited exercise stress test, a 6-minute walk test, or use of a Rating of Perceived Exertion scale. Prescribed exercise generally targets moderate activity in the range of 50% to 70% of peak estimated functional capacity. In the appropriate clinical context, high-functioning patients can be offered high-intensity interval training instead of moderate exercise, as they confer similar benefits.¹⁰

Risk-factor reduction. Comprehensive risk-factor reduction can address smoking, hypertension, high cholesterol, diabetes, obesity, and diet, as well as psychosocial issues such as stress, anxiety, depression, and alcohol use. Sexual activity counseling may also be included.

Education classes are aimed at helping patients understand cardiovascular disease and empowering them to manage their medical treatment and lifestyle modifications.⁶

Phase 3: Lifetime maintenance

In phase 3, patients independently continue risk-factor modification and physical activity without cardiac monitoring. Most cardiac rehabilitation programs offer transition-to-maintenance classes after completion of phase 2; this may be a welcome option, particularly for those who have developed a good routine and rapport with the staff and other participants. Others may opt for an independent program, using their own home equipment or a local health club.

EXERCISE: MOSTLY SAFE, WITH PROVEN BENEFITS

The safety of cardiac rehabilitation is well established, with a low risk of major cardiovascular complications. A US study in the early 1980s of 167 cardiac rehabilitation programs found 1 cardiac arrest for every 111,996 exercise hours, 1 myocardial infarction per 293,990 exercise hours, and 1 fatality per 783,972 exercise hours.¹¹ A 2006 study of more than 65 cardiac rehabilitation centers in France found 1 cardiac event per 8,484 exercise tests and 1.3 cardiac arrests per 1 million exercise hours.¹²

The benefits of cardiac rehabilitation are numerous and substantial.^9,13–17 A 2016 Cochrane review and meta-analysis of 63 randomized controlled trials with 14,486 participants found a reduced rate of cardiovascular mortality (relative risk [RR] 0.74, 95% confidence interval [CI] 0.64–0.86), with a number needed to treat of 37, and fewer hospital re­admissions (RR 0.82, 95% CI 0.70–0.96).⁹

Reductions in mortality rates are dose-dependent. A study of more than 30,000 Medicare beneficiaries who participated in cardiac rehabilitation found that those who attended more sessions had a lower rate of morbidity and death at 4 years, particularly if they participated in more than 11 sessions. Those who attended the full 36 sessions had a mortality rate 47% lower than those who attended a single session.¹⁷ There was a 15% reduction in mortality for those who attended 36 sessions compared with 24 sessions, a 28% lower risk with attending 36 sessions compared with 12. After adjustment, each additional 6 sessions was associated with a 6% reduction in mortality. The curves continued to separate up to 4 years.

The benefits of cardiac rehabilitation go beyond risk reduction and include improved functional capacity, greater ease with activities of daily living, and improved quality of life.⁹ Patients receive structure and support from the management team and other participants, which may provide an additional layer of friendship and psychosocial support for making lifestyle changes.

Is the overall mortality rate improved?

In the modern era, with access to optimal medical therapy and drug-eluting stents, one might expect only small additional benefit from cardiac rehabilitation. The 2016 Cochrane review and meta-analysis found that although cardiac rehabilitation contributed to improved cardiovascular mortality rates and health-related quality of life, no significant reduction was detected in the rate of death from all causes.⁸ But the analysis did not necessarily support removing the claim of reduced all-cause mortality for cardiac rehabilitation: only randomized controlled trials were examined, and the quality of evidence for each outcome was deemed to be low to moderate because of a general paucity of reports, including many small trials that followed patients for less than 12 months.

A large cohort analysis¹⁵ with more than 73,000 patients who had undergone cardiac rehabilitation found a relative reduction in mortality rate of 58% at 1 year and 21% to 34% at 5 years, with elderly women gaining the most benefit. In the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial, with more than 2,300 patients followed for a median of 2.5 years, exercise training for heart failure was associated with reduced rates of all-cause mortality or hospitalization (HR 0.89, 95% CI 0.81–0.99; P = .03) and of cardiovascular mortality or heart failure hospitalization (HR 0.85, 95% CI 0.74–0.99; P = .03).¹⁸

Regardless of the precise reduction in all-cause mortality, the cardiovascular and health-quality outcomes of cardiac rehabilitation clearly indicate benefit. More trials with follow-up longer than 1 year are needed to definitively determine the impact of cardiac rehabilitation on the all-cause mortality rate.

WHO SHOULD BE OFFERED CARDIAC REHABILITATION?

The 2006 CMS coverage criteria listed the indications for cardiac rehabilitation as myocardial infarction within the preceding 12 months, coronary artery bypass surgery, stable angina pectoris, heart valve repair or replacement, percutaneous coronary intervention, and heart or heart-lung transplant.

In 2014, stable chronic systolic heart failure was added to the list (Table 2). Qualifications include New York Heart Association class II (mild symptoms, slight limitation of activity) to class IV (severe limitations, symptoms at rest), an ejection fraction of 35% or less, and being stable on optimal medical therapy for at least 6 weeks.

In 2017, CMS approved supervised exercise therapy for peripheral arterial disease. Supervised exercise has a class 1 recommendation by the American Heart Association and American College of Cardiology for treating intermittent claudication. Supervised exercise therapy can increase walking distance by 180% and is superior to medical therapy alone. Unsupervised exercise has a class 2b recommendation.^19,20

Other patients may not qualify for phase 2 cardiac rehabilitation according to CMS or private insurance but could benefit from an exercise prescription and enrollment in a local phase 3 or home exercise program. Indications might include diabetes, obesity, metabolic syndrome, atrial fibrillation, postural orthostatic tachycardia syndrome, and nonalcoholic steatohepatitis. The benefits of cardiac rehabilitation after newer, less-invasive procedures for transcatheter valve repair and replacement are not well established, and more research is needed in this area.

WHEN TO REFER

Ades et al have defined cardiac rehabilitation referral as a combination of electronic medical records order, patient-physician discussion, and receipt of an order by a cardiac rehabilitation program.²¹

Ideally, referral for outpatient cardiac rehabilitation should take place at the time of hospital discharge. The AACVPR endorses a “cardiovascular continuum of care” model that emphasizes a smooth transition from inpatient to outpatient programs.⁶ Inpatient referral is a strong predictor of cardiac rehabilitation enrollment, and lack of referral in phase 1 negatively affects enrollment rates.

Depending on the diagnosis, US and Canadian guidelines recommend cardiac rehabilitation starting within 1 to 4 weeks of the index event, with acceptable wait times up to 60 days.^6,22 In the United Kingdom, referral is recommended within 24 hours of patient eligibility; assessment for a cardiovascular prevention and rehabilitation program, with a defined pathway and individual goals, is expected to be completed within 10 working days of referral.²³ Such a standard is difficult to meet in the United States, where the time from hospital discharge to cardiac rehabilitation program enrollment averages 35 days.^24,25

After an uncomplicated myocardial infarction or percutaneous coronary intervention, patients with a normal or mildly reduced left ventricular ejection fraction should start outpatient cardiac rehabilitation within 14 days of the index event. For such cases, cardiac rehabilitation has been shown to be safe within 1 to 2 weeks of hospital discharge and is associated with increased participation rates.

REHABILITATION IS STILL UNDERUSED

Despite its significant benefits, cardiac rehabilitation is underused for many reasons.

Referral rates vary

A study using the 1997 Medicare claims data­base showed national referral rates of only 14% after myocardial infarction and 31% after coronary artery bypass grafting.³¹

A later study using the National Cardiovascular Data Registry between 2009 and 2017 found that the situation had improved, with a referral rate of about 60% for patients undergoing percutaneous coronary intervention.³² Nevertheless, referral rates for cardiac rehabilitation remain highly variable and still lag behind other CMS quality measures for optimal medical therapy after acute myocardial infarction (Figure 1). Factors associated with higher referral rates included ST-segment elevation myocardial infarction, non-ST-segment elevation myocardial infarction, care in a high-volume center for percutaneous coronary intervention, and care in a private or community hospital in a Midwestern state. Small Midwestern hospitals generally had referral rates of over 80%, while major teaching hospitals and hospital systems on the East Coast and the West Coast had referral rates of less than 20%. Unlike some studies, this study found that insurance status had little bearing on referral rates.

Other studies found lower referral rates for women and patients with comorbidities such as previous coronary artery bypass grafting, diabetes, and heart failure.^33,34

In the United Kingdom, patients with heart failure made up only 5% of patients in cardiac rehabilitation; only 7% to 20% of patients with a heart failure diagnosis were referred to cardiac rehabilitation from general and cardiology wards.³⁵

Enrollment, completion rates even lower

Rates of referral for cardiac rehabilitation do not equate to rates of enrollment or participation. Enrollment was 50% in the United Kingdom in 2016.³⁵ A 2015 US study evaluated 58,269 older patients eligible for cardiac rehabilitation after acute myocardial infarction;  62% were referred for cardiac rehabilitation at the time of discharge, but only 23% of the total attended at least 1 session, and just 5% of the total completed 36 or more sessions.³⁶

BARRIERS, OPPORTUNITIES TO IMPROVE

The underuse of cardiac rehabilitation in the United States has led to an American Heart Association presidential advisory on the referral, enrollment, and delivery of cardiac rehabilitation.³⁴ Dozens of barriers are mentioned, with several standing out as having the largest impact: lack of physician referral, weak endorsement by the prescribing provider, female sex of patients, lack of program availability, work-related hardship, low socioeconomic status, and lack of or limited healthcare insurance. Copayments have also become a major barrier, often ranging from $20 to $40 per session for patients with Medicare.

The Million Hearts Initiative has established a goal of 70% cardiac rehabilitation compliance for eligible patients by 2022, a goal they estimate could save 25,000 lives and prevent 180,000 hospitalizations annually.²¹

Lack of physician awareness and lack of referral may be the most modifiable factors with the capacity to have the largest impact. Increasing physician awareness is a top priority not only for primary care providers, but also for cardiologists. In 2014, CMS made referral for cardiac rehabilitation a quality measure that is trackable and reportable. CMS has also proposed models that would incentivize participation by increasing reimbursement for services provided, but these models have been halted.

Additional efforts to increase cardiac rehabilitation referral and participation include automated order sets, increased caregiver education, and early morning or late evening classes, single-sex classes, home or mobile-based exercise programs, and parking and transportation assistance.³⁴ Grace et al³⁷ reported that referral rates rose to 86% when a cardiac rehabilitation order was integrated into the electronic medical record and combined with a hospital liaison to educate patients about their need for cardiac rehabilitation. Lowering patient copayments would also be a good idea. We have recently seen some creative ways to reduce copayments, including philanthropy and grants.

Cardiac rehabilitation has a class 1 indication (ie, strong recommendation) after heart surgery, myocardial infarction, or coronary intervention, and for stable angina or peripheral artery disease. It has a class 2a indication (ie, moderate recommendation) for stable systolic heart failure. Yet it is still under­utilized despite its demonstrated benefits, endorsement by most recognized cardiovascular societies, and coverage by the US Centers for Medicare and Medicaid Services (CMS).

Here, we review cardiac rehabilitation—its benefits, appropriate indications, barriers to referral and enrollment, and efforts to increase its use.

EXERCISE: SLOW TO BE ADOPTED

In 1772, William Heberden (also remembered today for describing swelling of the distal interphalangeal joints in osteoarthritis) described¹ a patient with angina pectoris who “set himself a task of sawing wood for half an hour every day, and was nearly cured.”

Despite early clues, it would be some time before the medical community would recognize the benefits of exercise for cardiovascular health. Before the 1930s, immobilization and extended bedrest were encouraged for up to 6 weeks after a cardiovascular event, leading to significant deconditioning.² Things slowly began to change in the 1940s with Levine’s introduction of up-to-chair therapy,³ and short daily walks were introduced in the 1950s. Over time, the link between a sedentary lifestyle and cardiovascular disease was studied and led to greater investigation into the benefits of exercise, propelling us into the modern era.^4,5

CARDIAC REHABILITATION: COMPREHENSIVE RISK REDUCTION

The American Association of Cardiovascular and Pulmonary Rehabilitation (AACVPR) defines cardiac rehabilitation as the provision of comprehensive long-term services involving medical evaluation, prescriptive exercise, cardiac risk-factor modification, education, counseling, and behavioral interventions.⁶ CMS defines it as a physician-supervised program that furnishes physician-prescribed exercise, cardiac risk-factor modification (including education, counseling, and behavioral intervention), psychosocial assessment, outcomes assessment, and other items and services.⁷

In general, most cardiac rehabilitation programs provide medically supervised exercise and patient education designed to improve cardiac health and functional status. Risk factors are targeted to reduce disability and rates of morbidity and mortality, to improve functional capacity, and to alleviate activity-related symptoms.

FROM HOSPITAL TO SELF-MAINTENANCE

Phase 1: Inpatient rehabilitation

Phase 1 typically takes place in the inpatient setting, often after open heart surgery (eg, coronary artery bypass grafting, valve repair or replacement, heart transplant), myocardial infarction, or percutaneous coronary intervention. This phase may last only a few days, especially in the current era of short hospital stays.

During phase 1, patients discuss their health situation and goals with their primary provider or cardiologist and receive education about recovery and cardiovascular risk factors. Early mobilization to prepare for discharge and to resume simple activities of daily living is emphasized. Depending on the institution, phase 1 exercise may involve simple ambulation on the ward or using equipment such as a stationary bike or treadmill.⁶ Phase 2 enrollment ideally is set up before discharge.

Phase 2: Limited-time outpatient rehabilitation

Phase 2 traditionally takes place in a hospital-based outpatient facility and consists of a physician-supervised multidisciplinary program. Growing evidence shows that home-based cardiac rehabilitation may be as effective as a medical facility-based program and should be an option for patients who have difficulty getting access to a traditional program.⁸

A phase 2 program takes a threefold approach, consisting of exercise, aggressive risk-factor modification, and education classes. A Cochrane review⁹ included programs that also incorporated behavioral modification and psychosocial support as a means of secondary prevention, underscoring the evolving definition of cardiac rehabilitation.

During the initial phase 2 visit, an individualized treatment plan is developed, incorporating an exercise prescription and realistic goals for secondary prevention. Sessions typically take place 3 times a week for up to 36 sessions; usually, options are available for less frequent weekly attendance for a longer period to achieve a full course. In some cases, patients may qualify for up to 72 sessions, particularly if they have not progressed as expected.

Exercise. As part of the initial evaluation, AACVPR guidelines⁶ suggest an exercise test­—eg, a symptom-limited exercise stress test, a 6-minute walk test, or use of a Rating of Perceived Exertion scale. Prescribed exercise generally targets moderate activity in the range of 50% to 70% of peak estimated functional capacity. In the appropriate clinical context, high-functioning patients can be offered high-intensity interval training instead of moderate exercise, as they confer similar benefits.¹⁰

Risk-factor reduction. Comprehensive risk-factor reduction can address smoking, hypertension, high cholesterol, diabetes, obesity, and diet, as well as psychosocial issues such as stress, anxiety, depression, and alcohol use. Sexual activity counseling may also be included.

Education classes are aimed at helping patients understand cardiovascular disease and empowering them to manage their medical treatment and lifestyle modifications.⁶

Phase 3: Lifetime maintenance

In phase 3, patients independently continue risk-factor modification and physical activity without cardiac monitoring. Most cardiac rehabilitation programs offer transition-to-maintenance classes after completion of phase 2; this may be a welcome option, particularly for those who have developed a good routine and rapport with the staff and other participants. Others may opt for an independent program, using their own home equipment or a local health club.

EXERCISE: MOSTLY SAFE, WITH PROVEN BENEFITS

The safety of cardiac rehabilitation is well established, with a low risk of major cardiovascular complications. A US study in the early 1980s of 167 cardiac rehabilitation programs found 1 cardiac arrest for every 111,996 exercise hours, 1 myocardial infarction per 293,990 exercise hours, and 1 fatality per 783,972 exercise hours.¹¹ A 2006 study of more than 65 cardiac rehabilitation centers in France found 1 cardiac event per 8,484 exercise tests and 1.3 cardiac arrests per 1 million exercise hours.¹²

The benefits of cardiac rehabilitation are numerous and substantial.^9,13–17 A 2016 Cochrane review and meta-analysis of 63 randomized controlled trials with 14,486 participants found a reduced rate of cardiovascular mortality (relative risk [RR] 0.74, 95% confidence interval [CI] 0.64–0.86), with a number needed to treat of 37, and fewer hospital re­admissions (RR 0.82, 95% CI 0.70–0.96).⁹

Reductions in mortality rates are dose-dependent. A study of more than 30,000 Medicare beneficiaries who participated in cardiac rehabilitation found that those who attended more sessions had a lower rate of morbidity and death at 4 years, particularly if they participated in more than 11 sessions. Those who attended the full 36 sessions had a mortality rate 47% lower than those who attended a single session.¹⁷ There was a 15% reduction in mortality for those who attended 36 sessions compared with 24 sessions, a 28% lower risk with attending 36 sessions compared with 12. After adjustment, each additional 6 sessions was associated with a 6% reduction in mortality. The curves continued to separate up to 4 years.

The benefits of cardiac rehabilitation go beyond risk reduction and include improved functional capacity, greater ease with activities of daily living, and improved quality of life.⁹ Patients receive structure and support from the management team and other participants, which may provide an additional layer of friendship and psychosocial support for making lifestyle changes.

Is the overall mortality rate improved?

In the modern era, with access to optimal medical therapy and drug-eluting stents, one might expect only small additional benefit from cardiac rehabilitation. The 2016 Cochrane review and meta-analysis found that although cardiac rehabilitation contributed to improved cardiovascular mortality rates and health-related quality of life, no significant reduction was detected in the rate of death from all causes.⁸ But the analysis did not necessarily support removing the claim of reduced all-cause mortality for cardiac rehabilitation: only randomized controlled trials were examined, and the quality of evidence for each outcome was deemed to be low to moderate because of a general paucity of reports, including many small trials that followed patients for less than 12 months.

A large cohort analysis¹⁵ with more than 73,000 patients who had undergone cardiac rehabilitation found a relative reduction in mortality rate of 58% at 1 year and 21% to 34% at 5 years, with elderly women gaining the most benefit. In the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial, with more than 2,300 patients followed for a median of 2.5 years, exercise training for heart failure was associated with reduced rates of all-cause mortality or hospitalization (HR 0.89, 95% CI 0.81–0.99; P = .03) and of cardiovascular mortality or heart failure hospitalization (HR 0.85, 95% CI 0.74–0.99; P = .03).¹⁸

Regardless of the precise reduction in all-cause mortality, the cardiovascular and health-quality outcomes of cardiac rehabilitation clearly indicate benefit. More trials with follow-up longer than 1 year are needed to definitively determine the impact of cardiac rehabilitation on the all-cause mortality rate.

WHO SHOULD BE OFFERED CARDIAC REHABILITATION?

The 2006 CMS coverage criteria listed the indications for cardiac rehabilitation as myocardial infarction within the preceding 12 months, coronary artery bypass surgery, stable angina pectoris, heart valve repair or replacement, percutaneous coronary intervention, and heart or heart-lung transplant.

In 2014, stable chronic systolic heart failure was added to the list (Table 2). Qualifications include New York Heart Association class II (mild symptoms, slight limitation of activity) to class IV (severe limitations, symptoms at rest), an ejection fraction of 35% or less, and being stable on optimal medical therapy for at least 6 weeks.

In 2017, CMS approved supervised exercise therapy for peripheral arterial disease. Supervised exercise has a class 1 recommendation by the American Heart Association and American College of Cardiology for treating intermittent claudication. Supervised exercise therapy can increase walking distance by 180% and is superior to medical therapy alone. Unsupervised exercise has a class 2b recommendation.^19,20

Other patients may not qualify for phase 2 cardiac rehabilitation according to CMS or private insurance but could benefit from an exercise prescription and enrollment in a local phase 3 or home exercise program. Indications might include diabetes, obesity, metabolic syndrome, atrial fibrillation, postural orthostatic tachycardia syndrome, and nonalcoholic steatohepatitis. The benefits of cardiac rehabilitation after newer, less-invasive procedures for transcatheter valve repair and replacement are not well established, and more research is needed in this area.

WHEN TO REFER

Ades et al have defined cardiac rehabilitation referral as a combination of electronic medical records order, patient-physician discussion, and receipt of an order by a cardiac rehabilitation program.²¹

Ideally, referral for outpatient cardiac rehabilitation should take place at the time of hospital discharge. The AACVPR endorses a “cardiovascular continuum of care” model that emphasizes a smooth transition from inpatient to outpatient programs.⁶ Inpatient referral is a strong predictor of cardiac rehabilitation enrollment, and lack of referral in phase 1 negatively affects enrollment rates.

Depending on the diagnosis, US and Canadian guidelines recommend cardiac rehabilitation starting within 1 to 4 weeks of the index event, with acceptable wait times up to 60 days.^6,22 In the United Kingdom, referral is recommended within 24 hours of patient eligibility; assessment for a cardiovascular prevention and rehabilitation program, with a defined pathway and individual goals, is expected to be completed within 10 working days of referral.²³ Such a standard is difficult to meet in the United States, where the time from hospital discharge to cardiac rehabilitation program enrollment averages 35 days.^24,25

After an uncomplicated myocardial infarction or percutaneous coronary intervention, patients with a normal or mildly reduced left ventricular ejection fraction should start outpatient cardiac rehabilitation within 14 days of the index event. For such cases, cardiac rehabilitation has been shown to be safe within 1 to 2 weeks of hospital discharge and is associated with increased participation rates.

REHABILITATION IS STILL UNDERUSED

Despite its significant benefits, cardiac rehabilitation is underused for many reasons.

Referral rates vary

A study using the 1997 Medicare claims data­base showed national referral rates of only 14% after myocardial infarction and 31% after coronary artery bypass grafting.³¹

A later study using the National Cardiovascular Data Registry between 2009 and 2017 found that the situation had improved, with a referral rate of about 60% for patients undergoing percutaneous coronary intervention.³² Nevertheless, referral rates for cardiac rehabilitation remain highly variable and still lag behind other CMS quality measures for optimal medical therapy after acute myocardial infarction (Figure 1). Factors associated with higher referral rates included ST-segment elevation myocardial infarction, non-ST-segment elevation myocardial infarction, care in a high-volume center for percutaneous coronary intervention, and care in a private or community hospital in a Midwestern state. Small Midwestern hospitals generally had referral rates of over 80%, while major teaching hospitals and hospital systems on the East Coast and the West Coast had referral rates of less than 20%. Unlike some studies, this study found that insurance status had little bearing on referral rates.

Other studies found lower referral rates for women and patients with comorbidities such as previous coronary artery bypass grafting, diabetes, and heart failure.^33,34

In the United Kingdom, patients with heart failure made up only 5% of patients in cardiac rehabilitation; only 7% to 20% of patients with a heart failure diagnosis were referred to cardiac rehabilitation from general and cardiology wards.³⁵

Enrollment, completion rates even lower

Rates of referral for cardiac rehabilitation do not equate to rates of enrollment or participation. Enrollment was 50% in the United Kingdom in 2016.³⁵ A 2015 US study evaluated 58,269 older patients eligible for cardiac rehabilitation after acute myocardial infarction;  62% were referred for cardiac rehabilitation at the time of discharge, but only 23% of the total attended at least 1 session, and just 5% of the total completed 36 or more sessions.³⁶

BARRIERS, OPPORTUNITIES TO IMPROVE

The underuse of cardiac rehabilitation in the United States has led to an American Heart Association presidential advisory on the referral, enrollment, and delivery of cardiac rehabilitation.³⁴ Dozens of barriers are mentioned, with several standing out as having the largest impact: lack of physician referral, weak endorsement by the prescribing provider, female sex of patients, lack of program availability, work-related hardship, low socioeconomic status, and lack of or limited healthcare insurance. Copayments have also become a major barrier, often ranging from $20 to $40 per session for patients with Medicare.

The Million Hearts Initiative has established a goal of 70% cardiac rehabilitation compliance for eligible patients by 2022, a goal they estimate could save 25,000 lives and prevent 180,000 hospitalizations annually.²¹

Lack of physician awareness and lack of referral may be the most modifiable factors with the capacity to have the largest impact. Increasing physician awareness is a top priority not only for primary care providers, but also for cardiologists. In 2014, CMS made referral for cardiac rehabilitation a quality measure that is trackable and reportable. CMS has also proposed models that would incentivize participation by increasing reimbursement for services provided, but these models have been halted.

Additional efforts to increase cardiac rehabilitation referral and participation include automated order sets, increased caregiver education, and early morning or late evening classes, single-sex classes, home or mobile-based exercise programs, and parking and transportation assistance.³⁴ Grace et al³⁷ reported that referral rates rose to 86% when a cardiac rehabilitation order was integrated into the electronic medical record and combined with a hospital liaison to educate patients about their need for cardiac rehabilitation. Lowering patient copayments would also be a good idea. We have recently seen some creative ways to reduce copayments, including philanthropy and grants.

Chronic kidney disease (CKD) is estimated to affect 14% of Americans, but it is likely underdiagnosed because it is often asymptomatic.^1,2 Its prevalence is even higher in patients who undergo surgery—up to 30% in cardiac surgery.³ Its impact on surgical outcomes is substantial.⁴ Importantly, patients with CKD are at higher risk of postoperative acute kidney injury (AKI), which is also associated with adverse outcomes. Thus, it is important to recognize, assess, and manage abnormal renal function in surgical patients.

WHAT IS THE IMPACT ON POSTOPERATIVE OUTCOMES?

Cardiac surgery outcomes

Moreover, in patients undergoing coronary artery bypass grafting (CABG), the worse the renal dysfunction, the higher the long-term mortality rate. Patients with moderate (stage 3) CKD had a 3.5 times higher odds of in-hospital mortality compared with patients with normal renal function, rising to 8.8 with severe (stage 4) and to 9.6 with dialysis-dependent (stage 5) CKD.¹¹

The mechanisms linking CKD with negative cardiac outcomes are unclear, but many possibilities exist. CKD is an independent risk factor for coronary artery disease and shares underlying risk factors such as hypertension and diabetes. Cardiac surgery patients with CKD are also more likely to have diabetes, left ventricular dysfunction, and peripheral vascular disease.

Noncardiac surgery outcomes

CKD is also associated with adverse outcomes in noncardiac surgery patients, especially at higher levels of renal dysfunction.^12–14 For example, in patients who underwent major noncardiac surgery, compared with patients in stage 1 (estimated GFR > 90 mL/min/1.73 m²), the odds ratios for all-cause mortality were as follows:

• 0.8 for patients with stage 2 CKD
• 2.2 in stage 3a
• 2.8 in stage 3b
• 11.3 in stage 4
• 5.8 in stage 5.¹⁴

The association between estimated GFR and all-cause mortality was not statistically significant (P = .071), but statistically significant associations were observed between estimated GFR and major adverse cardiovascular events (P < .001) and hospital length of stay (P < .001).

The association of CKD with major adverse outcomes and death in both cardiac and noncardiac surgical patients demonstrates the importance of understanding this risk, identifying patients with CKD preoperatively, and taking steps to lower the risk.

WHAT IS THE IMPACT OF ACUTE KIDNEY INJURY?

AKI is a common and serious complication of surgery, especially cardiac surgery. It has been associated with higher rates of morbidity, mortality, and cardiovascular events, longer hospital length of stay, and higher cost.

Several groups have proposed criteria for defining AKI and its severity; the KDIGO criteria are the most widely accepted.¹⁵ These define AKI as an increase in serum creatinine concentration of 0.3 mg/dL or more within 48 hours or at least 1.5 times the baseline value within 7 days, or urine volume less than 0.5 mL/kg/hour for more than 6 hours. There are 3 stages of severity:

• Stage 1—an increase in serum creatinine of 1.5 to 1.9 times baseline, an absolute increase of at least 0.3 mg/dL, or urine output less than 0.5 mL/kg/hour for 6 to 12 hours
• Stage 2—an increase in serum creatinine of 2.0 to 2.9 times baseline or urine output less than 0.5 mmL/kg/hour for 12 or more hours
• Stage 3—an increase in serum creatinine of 3 times baseline, an absolute increase of at least 4 mg/dL, initiation of renal replacement therapy, urine output less than 0.3 mL/kg/hour for 24 or more hours, or anuria for 12 or more hours.¹⁵

Multiple factors associated with surgery may contribute to AKI, including hemodynamic instability, volume shifts, blood loss, use of heart-lung bypass, new medications, activation of the inflammatory cascade, oxidative stress, and anemia.

AKI in cardiac surgery

The incidence of AKI is high in cardiac surgery. In a meta-analysis of 46 studies (N = 242,000), its incidence in cardiopulmonary bypass surgery was about 18%, with 2.1% of patients needing renal replacement therapy.¹⁶ However, the incidence varied considerably from study to study, ranging from 1% to 53%, and was influenced by the definition of AKI, the type of cardiac surgery, and the patient population.¹⁶

Cardiac surgery-associated AKI adversely affects outcomes. Several studies have shown that cardiac surgery patients who develop AKI have higher rates of death and stroke.^16–21 More severe AKI confers higher mortality rates, with the highest mortality rate in patients who need renal replacement therapy, approximately 37%.¹⁷ Patients with cardiac surgery-associated AKI also have a longer hospital length of stay and significantly higher costs of care.^17,18

Long-term outcomes are also negatively affected by AKI. In cardiac surgery patients with AKI who had completely recovered renal function by the time they left the hospital, the 2-year incidence rate of CKD was 6.8%, significantly higher than the 0.2% rate in patients who did not develop AKI.¹⁹ The 2-year survival rates also were significantly worse for patients who developed postoperative AKI (82.3% vs 93.7%). Similarly, in patients undergoing CABG who had normal renal function before surgery, those who developed AKI postoperatively had significantly shorter long-term survival rates.²⁰ The effect does not require a large change in renal function. An increase in creatinine as small as 0.3 mg/dL has been associated with a higher rate of death and a long-term risk of end-stage renal disease that is 3 times higher.²¹

Cardiac surgery

CKD is a risk factor not only after cardiac surgery but also after percutaneous procedures. In a meta-analysis of 4,992 patients with CKD who underwent transcatheter aortic valve replacement, both moderate and severe CKD increased the odds of AKI, early stroke, the need for dialysis, and all-cause and cardiovascular mortality at 1 year.^22,23 Increased rates of AKI also have been found in patients with CKD undergoing CABG surgery.²⁴ These results point to a synergistic effect between AKI and CKD, with outcomes much worse in combination than alone.

In cardiac surgery, the most important patient risk factors associated with a higher incidence of postoperative AKI are age older than 75, CKD, preoperative heart failure, and prior myocardial infarction.^19,25 Diabetes is an additional independent risk factor, with type 1 conferring higher risk than type 2.²⁶ Preoperative use of angiotensin-converting enzyme (ACE) inhibitors may or may not be a risk factor for cardiac surgery-associated AKI, with some studies finding increased risk and others finding reduced rates.^27,28

Anemia, which may be related to either patient or surgical risk factors (eg, intraoperative blood loss), also increases the risk of AKI in cardiac surgery.^29,30 A retrospective study of CABG surgery patients found that intraoperative hemoglobin levels below 8 g/dL were associated with a 25% to 30% incidence of AKI, compared with 15% to 20% with hemoglobin levels above 9 g/dL.²⁹ Additionally, having severe hypotension (mean arterial pressure < 50 mm Hg) significantly increased the AKI rates in the low-hemoglobin group.²⁹ Similar results were reported in a later study.³⁰

Among surgical factors, several randomized controlled trials have shown that off-pump CABG is associated with a significantly lower risk of postoperative AKI than on-pump CABG; however, this difference did not translate into any long-term difference in mortality rates.^31,32 Longer cardiopulmonary bypass time is strongly associated with a higher incidence of AKI and postoperative death.³³

Noncardiac surgery

AKI is less common after noncardiac surgery; however, outcomes are severe in patients in whom it occurs. In a study of 15,102 noncardiac surgery patients, only 0.8% developed AKI and 0.1% required renal replacement therapy.³⁴

Risk factors after noncardiac surgery are similar to those after cardiac surgery (Table 3).^34–36 Factors with the greatest impact are older age, peripheral vascular occlusive disease, chronic obstructive pulmonary disease necessitating chronic bronchodilator therapy, high-risk surgery, hepatic disease, emergent or urgent surgery, and high body mass index.

Surgical risk factors include total vasopressor dose administered, use of a vasopressor infusion, and diuretic administration.³⁴ In addition, intraoperative hypotension is associated with a higher risk of AKI, major adverse cardiac events, and 30-day mortality.³⁷

Noncardiac surgery patients with postoperative AKI have significantly higher rates of 30-day readmissions, 1-year progression to end-stage renal disease, and mortality than patients who do not develop AKI.³⁵ Additionally, patients with AKI have significantly higher rates of cardiovascular complications (33.3% vs 11.3%) and death (6.1% vs 0.9%), as well as a significantly longer length of hospital stay.^34,36

CAN WE DECREASE THE IMPACT OF RENAL DISEASE IN SURGERY?

Before surgery, practitioners need to identify patients at risk of AKI, implement possible risk-reduction measures, and, afterward, treat it early in its course if it occurs.

The preoperative visit is the ideal time to assess a patient’s risk of postoperative renal dysfunction. Laboratory tests can identify risks based on surgery type, age, hypertension, the presence of CKD, and medications that affect renal function. However, the basic chemistry panel is abnormal in only 8.2% of patients and affects management in just 2.6%, requiring the clinician to target testing to patients at high risk.³⁸

Patients with a significant degree of renal dysfunction, particularly those previously undiagnosed, may benefit from additional preoperative testing and medication management. Perioperative management of medications that could adversely affect renal function should be carefully considered during the preoperative visit. In addition, the postoperative inpatient team needs to be informed about potentially nephrotoxic medications and medications that are renally cleared. Attention needs to be given to the renal impact of common perioperative medications such as nonsteroidal anti-inflammatory drugs, antibiotics, intravenous contrast, low-molecular-weight heparins, diuretics, ACE inhibitors, and angiotensin II receptor blockers. With the emphasis on opioid-sparing analgesics, it is particularly important to assess the risk of AKI if nonsteroidal anti-inflammatory drugs are part of the pain control plan.

Nephrology referral may help, especially for patients with a GFR less than 45 mL/min. This information enables more informed decision-making regarding the risks of adverse outcomes related to kidney disease.

WHAT TOOLS DO WE HAVE TO DIAGNOSE RENAL INJURY?

Several risk-prediction models have been developed to assess the postoperative risk of AKI in both cardiac and major noncardiac surgery patients. Although these models can identify risk factors, their clinical accuracy and utility have been questioned.

Biomarkers

Early diagnosis is the first step in managing AKI, allowing time to implement measures to minimize its impact.

Serum creatinine testing is widely used to measure renal function and diagnose AKI; however, it does not detect small reductions in renal function, and there is a time lag between renal insult and a rise in creatinine. The result is a delay to diagnosis of AKI.

Biomarkers other than creatinine have been studied for early detection of intraoperative and postoperative renal insult. These novel renal injury markers include the following:

Neutrophil gelatinase-associated lipocalin (NGAL). Two studies looked at plasma NGAL as an early marker of AKI in patients with CKD who were undergoing cardiac surgery.^39,40 One study found that by using NGAL instead of creatinine, postoperative AKI could be diagnosed an average of 20 hours earlier.³⁹ In addition, NGAL helped detect renal recovery earlier than creatinine.⁴⁰ The diagnostic cut-off values of NGAL were different for patients with CKD than for those without CKD.^39,40

Other novel markers include:

• Kidney injury marker 1
• N-acetyl-beta-D-glucosaminidase
• Cysteine C.

Although these biomarkers show some ability to detect renal injury, they provide only modest discrimination and are not widely available for clinical use.⁴¹ Current evidence does not support routine use of these markers in clinical settings.

Interventions to prevent or ameliorate the impact of CKD and AKI on surgical outcomes have been studied most extensively in cardiac surgery patients.

Aspirin. A retrospective study of 3,585 cardiac surgery patients with CKD found that preoperative aspirin use significantly lowered the incidence of postoperative AKI and 30-day mortality compared with patients not using aspirin.⁴² Aspirin use reduced 30-day mortality in CKD stages 1, 2, and 3 by 23.3%, 58%, and 70%, respectively. On the other hand, in the Perioperative Ischemic Evaluation (POISE) trial, in noncardiac surgery patients, neither aspirin nor clonidine started 2 to 4 hours preoperatively and continued up to 30 days after surgery altered the risk of AKI significantly more than placebo.⁴³

Statins have been ineffective in reducing the incidence of AKI in cardiac surgery patients. In fact, a meta-analysis of 8 interventional trials found an increased incidence of AKI in patients in whom statins were started perioperatively.⁴⁴ Erythropoietin was also found to be ineffective in the prevention of perioperative AKI in cardiac surgery patients in a separate study.⁴⁵

The evidence regarding other therapies has also varied.

N-acetylcysteine in high doses reduced the incidence of AKI in patients with CKD stage 3 and 4 undergoing CABG.⁴⁶ Another meta-analysis of 10 studies in cardiac surgery patients published recently did not show any benefit of N-acetylcysteine in reducing AKI.⁴⁷

Human atrial natriuretic peptide, given preoperatively to patients with CKD, reduced the acute and long-term creatinine rise as well as the number of cardiac events after CABG; however, it did not reduce mortality rates.⁴⁸

Renin-angiotensin system inhibitors, given preoperatively to patients with heart failure was associated with a decrease in the incidence of AKI in 1 study.⁴⁹

Dexmedetomidine is a highly selective alpha 2 adrenoreceptor agonist. A recent meta-analysis of 10 clinical trials found it beneficial in reducing the risk of perioperative AKI in cardiac surgery patients.⁵⁰ An earlier meta-analysis had similar results.⁵¹

Levosimendan is an inotropic vasodilator that improves cardiac output and renal perfusion in patients with systolic heart failure, and it has been hypothesized to decrease the risk of AKI after cardiac surgery. Previous data demonstrated that this drug reduced AKI and mortality; however, analysis was limited by small sample size and varying definitions of AKI.⁵² A recent meta-analysis showed that levosimendan was associated with a lower incidence of AKI but was also associated with an increased incidence of atrial fibrillation and no reduction in 30-day mortality.⁵³

Remote ischemic preconditioning is a procedure that subjects the kidneys to brief episodes of ischemia before surgery, protecting them when they are later subjected to prolonged ischemia or reperfusion injury. It has shown initial promising results in preventing AKI. In a randomized controlled trial in 240 patients at high risk of AKI, those who received remote ischemic preconditioning had an AKI incidence of 37.5% compared with 52.5% for controls (P = .02); however, the mortality rate was the same.⁵⁴ Similarly, remote ischemic preconditioning significantly lowered the incidence of AKI in nondiabetic patients undergoing CABG surgery compared with controls.⁵⁵

Fluid management. Renal perfusion is intimately related to the development of AKI, and there is evidence that both hypovolemia and excessive fluid resuscitation can increase the risk of AKI in noncardiac surgery patients.⁵⁶ Because of this, fluid management has also received attention in perioperative AKI. Goal-directed fluid management has been evaluated in noncardiac surgery patients, and it did not show any benefit in preventing AKI.⁵⁷ However, in a more recent retrospective study, postoperative positive fluid balance was associated with increased incidence of AKI compared with zero fluid balance. Negative fluid balance did not appear to have a detrimental effect.⁵⁸

RECOMMENDATIONS

No prophylactic therapy has yet been shown to definitively decrease the risk of postoperative AKI in all patients. Nevertheless, it is important to identify patients at risk during the preoperative visit, especially those with CKD. Many patients undergoing surgery have CKD, placing them at high risk of developing AKI in the perioperative period. The risk is particularly high with cardiac surgery.

Serum creatinine and urine output should be closely monitored perioperatively in at-risk patients. If AKI is diagnosed, practitioners need to identify and ameliorate the cause as early as possible.

Recommendations from KDIGO for perioperative prevention and management of AKI are listed in Table 4.¹⁵ These include avoiding additional nephrotoxic medications and adjusting the doses of renally cleared medications. Also, some patients may benefit from preoperative counseling and specialist referral.

Chronic kidney disease (CKD) is estimated to affect 14% of Americans, but it is likely underdiagnosed because it is often asymptomatic.^1,2 Its prevalence is even higher in patients who undergo surgery—up to 30% in cardiac surgery.³ Its impact on surgical outcomes is substantial.⁴ Importantly, patients with CKD are at higher risk of postoperative acute kidney injury (AKI), which is also associated with adverse outcomes. Thus, it is important to recognize, assess, and manage abnormal renal function in surgical patients.

WHAT IS THE IMPACT ON POSTOPERATIVE OUTCOMES?

Cardiac surgery outcomes

Moreover, in patients undergoing coronary artery bypass grafting (CABG), the worse the renal dysfunction, the higher the long-term mortality rate. Patients with moderate (stage 3) CKD had a 3.5 times higher odds of in-hospital mortality compared with patients with normal renal function, rising to 8.8 with severe (stage 4) and to 9.6 with dialysis-dependent (stage 5) CKD.¹¹

The mechanisms linking CKD with negative cardiac outcomes are unclear, but many possibilities exist. CKD is an independent risk factor for coronary artery disease and shares underlying risk factors such as hypertension and diabetes. Cardiac surgery patients with CKD are also more likely to have diabetes, left ventricular dysfunction, and peripheral vascular disease.

Noncardiac surgery outcomes

CKD is also associated with adverse outcomes in noncardiac surgery patients, especially at higher levels of renal dysfunction.^12–14 For example, in patients who underwent major noncardiac surgery, compared with patients in stage 1 (estimated GFR > 90 mL/min/1.73 m²), the odds ratios for all-cause mortality were as follows:

• 0.8 for patients with stage 2 CKD
• 2.2 in stage 3a
• 2.8 in stage 3b
• 11.3 in stage 4
• 5.8 in stage 5.¹⁴

The association between estimated GFR and all-cause mortality was not statistically significant (P = .071), but statistically significant associations were observed between estimated GFR and major adverse cardiovascular events (P < .001) and hospital length of stay (P < .001).

The association of CKD with major adverse outcomes and death in both cardiac and noncardiac surgical patients demonstrates the importance of understanding this risk, identifying patients with CKD preoperatively, and taking steps to lower the risk.

WHAT IS THE IMPACT OF ACUTE KIDNEY INJURY?

AKI is a common and serious complication of surgery, especially cardiac surgery. It has been associated with higher rates of morbidity, mortality, and cardiovascular events, longer hospital length of stay, and higher cost.

Several groups have proposed criteria for defining AKI and its severity; the KDIGO criteria are the most widely accepted.¹⁵ These define AKI as an increase in serum creatinine concentration of 0.3 mg/dL or more within 48 hours or at least 1.5 times the baseline value within 7 days, or urine volume less than 0.5 mL/kg/hour for more than 6 hours. There are 3 stages of severity:

• Stage 1—an increase in serum creatinine of 1.5 to 1.9 times baseline, an absolute increase of at least 0.3 mg/dL, or urine output less than 0.5 mL/kg/hour for 6 to 12 hours
• Stage 2—an increase in serum creatinine of 2.0 to 2.9 times baseline or urine output less than 0.5 mmL/kg/hour for 12 or more hours
• Stage 3—an increase in serum creatinine of 3 times baseline, an absolute increase of at least 4 mg/dL, initiation of renal replacement therapy, urine output less than 0.3 mL/kg/hour for 24 or more hours, or anuria for 12 or more hours.¹⁵

Multiple factors associated with surgery may contribute to AKI, including hemodynamic instability, volume shifts, blood loss, use of heart-lung bypass, new medications, activation of the inflammatory cascade, oxidative stress, and anemia.

AKI in cardiac surgery

The incidence of AKI is high in cardiac surgery. In a meta-analysis of 46 studies (N = 242,000), its incidence in cardiopulmonary bypass surgery was about 18%, with 2.1% of patients needing renal replacement therapy.¹⁶ However, the incidence varied considerably from study to study, ranging from 1% to 53%, and was influenced by the definition of AKI, the type of cardiac surgery, and the patient population.¹⁶

Cardiac surgery-associated AKI adversely affects outcomes. Several studies have shown that cardiac surgery patients who develop AKI have higher rates of death and stroke.^16–21 More severe AKI confers higher mortality rates, with the highest mortality rate in patients who need renal replacement therapy, approximately 37%.¹⁷ Patients with cardiac surgery-associated AKI also have a longer hospital length of stay and significantly higher costs of care.^17,18

Long-term outcomes are also negatively affected by AKI. In cardiac surgery patients with AKI who had completely recovered renal function by the time they left the hospital, the 2-year incidence rate of CKD was 6.8%, significantly higher than the 0.2% rate in patients who did not develop AKI.¹⁹ The 2-year survival rates also were significantly worse for patients who developed postoperative AKI (82.3% vs 93.7%). Similarly, in patients undergoing CABG who had normal renal function before surgery, those who developed AKI postoperatively had significantly shorter long-term survival rates.²⁰ The effect does not require a large change in renal function. An increase in creatinine as small as 0.3 mg/dL has been associated with a higher rate of death and a long-term risk of end-stage renal disease that is 3 times higher.²¹

Cardiac surgery

CKD is a risk factor not only after cardiac surgery but also after percutaneous procedures. In a meta-analysis of 4,992 patients with CKD who underwent transcatheter aortic valve replacement, both moderate and severe CKD increased the odds of AKI, early stroke, the need for dialysis, and all-cause and cardiovascular mortality at 1 year.^22,23 Increased rates of AKI also have been found in patients with CKD undergoing CABG surgery.²⁴ These results point to a synergistic effect between AKI and CKD, with outcomes much worse in combination than alone.

In cardiac surgery, the most important patient risk factors associated with a higher incidence of postoperative AKI are age older than 75, CKD, preoperative heart failure, and prior myocardial infarction.^19,25 Diabetes is an additional independent risk factor, with type 1 conferring higher risk than type 2.²⁶ Preoperative use of angiotensin-converting enzyme (ACE) inhibitors may or may not be a risk factor for cardiac surgery-associated AKI, with some studies finding increased risk and others finding reduced rates.^27,28

Anemia, which may be related to either patient or surgical risk factors (eg, intraoperative blood loss), also increases the risk of AKI in cardiac surgery.^29,30 A retrospective study of CABG surgery patients found that intraoperative hemoglobin levels below 8 g/dL were associated with a 25% to 30% incidence of AKI, compared with 15% to 20% with hemoglobin levels above 9 g/dL.²⁹ Additionally, having severe hypotension (mean arterial pressure < 50 mm Hg) significantly increased the AKI rates in the low-hemoglobin group.²⁹ Similar results were reported in a later study.³⁰

Among surgical factors, several randomized controlled trials have shown that off-pump CABG is associated with a significantly lower risk of postoperative AKI than on-pump CABG; however, this difference did not translate into any long-term difference in mortality rates.^31,32 Longer cardiopulmonary bypass time is strongly associated with a higher incidence of AKI and postoperative death.³³

Noncardiac surgery

AKI is less common after noncardiac surgery; however, outcomes are severe in patients in whom it occurs. In a study of 15,102 noncardiac surgery patients, only 0.8% developed AKI and 0.1% required renal replacement therapy.³⁴

Risk factors after noncardiac surgery are similar to those after cardiac surgery (Table 3).^34–36 Factors with the greatest impact are older age, peripheral vascular occlusive disease, chronic obstructive pulmonary disease necessitating chronic bronchodilator therapy, high-risk surgery, hepatic disease, emergent or urgent surgery, and high body mass index.

Surgical risk factors include total vasopressor dose administered, use of a vasopressor infusion, and diuretic administration.³⁴ In addition, intraoperative hypotension is associated with a higher risk of AKI, major adverse cardiac events, and 30-day mortality.³⁷

Noncardiac surgery patients with postoperative AKI have significantly higher rates of 30-day readmissions, 1-year progression to end-stage renal disease, and mortality than patients who do not develop AKI.³⁵ Additionally, patients with AKI have significantly higher rates of cardiovascular complications (33.3% vs 11.3%) and death (6.1% vs 0.9%), as well as a significantly longer length of hospital stay.^34,36

CAN WE DECREASE THE IMPACT OF RENAL DISEASE IN SURGERY?

Before surgery, practitioners need to identify patients at risk of AKI, implement possible risk-reduction measures, and, afterward, treat it early in its course if it occurs.

The preoperative visit is the ideal time to assess a patient’s risk of postoperative renal dysfunction. Laboratory tests can identify risks based on surgery type, age, hypertension, the presence of CKD, and medications that affect renal function. However, the basic chemistry panel is abnormal in only 8.2% of patients and affects management in just 2.6%, requiring the clinician to target testing to patients at high risk.³⁸

Patients with a significant degree of renal dysfunction, particularly those previously undiagnosed, may benefit from additional preoperative testing and medication management. Perioperative management of medications that could adversely affect renal function should be carefully considered during the preoperative visit. In addition, the postoperative inpatient team needs to be informed about potentially nephrotoxic medications and medications that are renally cleared. Attention needs to be given to the renal impact of common perioperative medications such as nonsteroidal anti-inflammatory drugs, antibiotics, intravenous contrast, low-molecular-weight heparins, diuretics, ACE inhibitors, and angiotensin II receptor blockers. With the emphasis on opioid-sparing analgesics, it is particularly important to assess the risk of AKI if nonsteroidal anti-inflammatory drugs are part of the pain control plan.

Nephrology referral may help, especially for patients with a GFR less than 45 mL/min. This information enables more informed decision-making regarding the risks of adverse outcomes related to kidney disease.

WHAT TOOLS DO WE HAVE TO DIAGNOSE RENAL INJURY?

Several risk-prediction models have been developed to assess the postoperative risk of AKI in both cardiac and major noncardiac surgery patients. Although these models can identify risk factors, their clinical accuracy and utility have been questioned.

Biomarkers

Early diagnosis is the first step in managing AKI, allowing time to implement measures to minimize its impact.

Serum creatinine testing is widely used to measure renal function and diagnose AKI; however, it does not detect small reductions in renal function, and there is a time lag between renal insult and a rise in creatinine. The result is a delay to diagnosis of AKI.

Biomarkers other than creatinine have been studied for early detection of intraoperative and postoperative renal insult. These novel renal injury markers include the following:

Neutrophil gelatinase-associated lipocalin (NGAL). Two studies looked at plasma NGAL as an early marker of AKI in patients with CKD who were undergoing cardiac surgery.^39,40 One study found that by using NGAL instead of creatinine, postoperative AKI could be diagnosed an average of 20 hours earlier.³⁹ In addition, NGAL helped detect renal recovery earlier than creatinine.⁴⁰ The diagnostic cut-off values of NGAL were different for patients with CKD than for those without CKD.^39,40

Other novel markers include:

• Kidney injury marker 1
• N-acetyl-beta-D-glucosaminidase
• Cysteine C.

Although these biomarkers show some ability to detect renal injury, they provide only modest discrimination and are not widely available for clinical use.⁴¹ Current evidence does not support routine use of these markers in clinical settings.

Interventions to prevent or ameliorate the impact of CKD and AKI on surgical outcomes have been studied most extensively in cardiac surgery patients.

Aspirin. A retrospective study of 3,585 cardiac surgery patients with CKD found that preoperative aspirin use significantly lowered the incidence of postoperative AKI and 30-day mortality compared with patients not using aspirin.⁴² Aspirin use reduced 30-day mortality in CKD stages 1, 2, and 3 by 23.3%, 58%, and 70%, respectively. On the other hand, in the Perioperative Ischemic Evaluation (POISE) trial, in noncardiac surgery patients, neither aspirin nor clonidine started 2 to 4 hours preoperatively and continued up to 30 days after surgery altered the risk of AKI significantly more than placebo.⁴³

Statins have been ineffective in reducing the incidence of AKI in cardiac surgery patients. In fact, a meta-analysis of 8 interventional trials found an increased incidence of AKI in patients in whom statins were started perioperatively.⁴⁴ Erythropoietin was also found to be ineffective in the prevention of perioperative AKI in cardiac surgery patients in a separate study.⁴⁵

The evidence regarding other therapies has also varied.

N-acetylcysteine in high doses reduced the incidence of AKI in patients with CKD stage 3 and 4 undergoing CABG.⁴⁶ Another meta-analysis of 10 studies in cardiac surgery patients published recently did not show any benefit of N-acetylcysteine in reducing AKI.⁴⁷

Human atrial natriuretic peptide, given preoperatively to patients with CKD, reduced the acute and long-term creatinine rise as well as the number of cardiac events after CABG; however, it did not reduce mortality rates.⁴⁸

Renin-angiotensin system inhibitors, given preoperatively to patients with heart failure was associated with a decrease in the incidence of AKI in 1 study.⁴⁹

Dexmedetomidine is a highly selective alpha 2 adrenoreceptor agonist. A recent meta-analysis of 10 clinical trials found it beneficial in reducing the risk of perioperative AKI in cardiac surgery patients.⁵⁰ An earlier meta-analysis had similar results.⁵¹

Levosimendan is an inotropic vasodilator that improves cardiac output and renal perfusion in patients with systolic heart failure, and it has been hypothesized to decrease the risk of AKI after cardiac surgery. Previous data demonstrated that this drug reduced AKI and mortality; however, analysis was limited by small sample size and varying definitions of AKI.⁵² A recent meta-analysis showed that levosimendan was associated with a lower incidence of AKI but was also associated with an increased incidence of atrial fibrillation and no reduction in 30-day mortality.⁵³

Remote ischemic preconditioning is a procedure that subjects the kidneys to brief episodes of ischemia before surgery, protecting them when they are later subjected to prolonged ischemia or reperfusion injury. It has shown initial promising results in preventing AKI. In a randomized controlled trial in 240 patients at high risk of AKI, those who received remote ischemic preconditioning had an AKI incidence of 37.5% compared with 52.5% for controls (P = .02); however, the mortality rate was the same.⁵⁴ Similarly, remote ischemic preconditioning significantly lowered the incidence of AKI in nondiabetic patients undergoing CABG surgery compared with controls.⁵⁵

Fluid management. Renal perfusion is intimately related to the development of AKI, and there is evidence that both hypovolemia and excessive fluid resuscitation can increase the risk of AKI in noncardiac surgery patients.⁵⁶ Because of this, fluid management has also received attention in perioperative AKI. Goal-directed fluid management has been evaluated in noncardiac surgery patients, and it did not show any benefit in preventing AKI.⁵⁷ However, in a more recent retrospective study, postoperative positive fluid balance was associated with increased incidence of AKI compared with zero fluid balance. Negative fluid balance did not appear to have a detrimental effect.⁵⁸

RECOMMENDATIONS

No prophylactic therapy has yet been shown to definitively decrease the risk of postoperative AKI in all patients. Nevertheless, it is important to identify patients at risk during the preoperative visit, especially those with CKD. Many patients undergoing surgery have CKD, placing them at high risk of developing AKI in the perioperative period. The risk is particularly high with cardiac surgery.

Serum creatinine and urine output should be closely monitored perioperatively in at-risk patients. If AKI is diagnosed, practitioners need to identify and ameliorate the cause as early as possible.

Recommendations from KDIGO for perioperative prevention and management of AKI are listed in Table 4.¹⁵ These include avoiding additional nephrotoxic medications and adjusting the doses of renally cleared medications. Also, some patients may benefit from preoperative counseling and specialist referral.

Chronic kidney disease (CKD) is estimated to affect 14% of Americans, but it is likely underdiagnosed because it is often asymptomatic.^1,2 Its prevalence is even higher in patients who undergo surgery—up to 30% in cardiac surgery.³ Its impact on surgical outcomes is substantial.⁴ Importantly, patients with CKD are at higher risk of postoperative acute kidney injury (AKI), which is also associated with adverse outcomes. Thus, it is important to recognize, assess, and manage abnormal renal function in surgical patients.

WHAT IS THE IMPACT ON POSTOPERATIVE OUTCOMES?

Cardiac surgery outcomes

Moreover, in patients undergoing coronary artery bypass grafting (CABG), the worse the renal dysfunction, the higher the long-term mortality rate. Patients with moderate (stage 3) CKD had a 3.5 times higher odds of in-hospital mortality compared with patients with normal renal function, rising to 8.8 with severe (stage 4) and to 9.6 with dialysis-dependent (stage 5) CKD.¹¹

The mechanisms linking CKD with negative cardiac outcomes are unclear, but many possibilities exist. CKD is an independent risk factor for coronary artery disease and shares underlying risk factors such as hypertension and diabetes. Cardiac surgery patients with CKD are also more likely to have diabetes, left ventricular dysfunction, and peripheral vascular disease.

Noncardiac surgery outcomes

CKD is also associated with adverse outcomes in noncardiac surgery patients, especially at higher levels of renal dysfunction.^12–14 For example, in patients who underwent major noncardiac surgery, compared with patients in stage 1 (estimated GFR > 90 mL/min/1.73 m²), the odds ratios for all-cause mortality were as follows:

• 0.8 for patients with stage 2 CKD
• 2.2 in stage 3a
• 2.8 in stage 3b
• 11.3 in stage 4
• 5.8 in stage 5.¹⁴

The association between estimated GFR and all-cause mortality was not statistically significant (P = .071), but statistically significant associations were observed between estimated GFR and major adverse cardiovascular events (P < .001) and hospital length of stay (P < .001).

The association of CKD with major adverse outcomes and death in both cardiac and noncardiac surgical patients demonstrates the importance of understanding this risk, identifying patients with CKD preoperatively, and taking steps to lower the risk.

WHAT IS THE IMPACT OF ACUTE KIDNEY INJURY?

AKI is a common and serious complication of surgery, especially cardiac surgery. It has been associated with higher rates of morbidity, mortality, and cardiovascular events, longer hospital length of stay, and higher cost.

Several groups have proposed criteria for defining AKI and its severity; the KDIGO criteria are the most widely accepted.¹⁵ These define AKI as an increase in serum creatinine concentration of 0.3 mg/dL or more within 48 hours or at least 1.5 times the baseline value within 7 days, or urine volume less than 0.5 mL/kg/hour for more than 6 hours. There are 3 stages of severity:

• Stage 1—an increase in serum creatinine of 1.5 to 1.9 times baseline, an absolute increase of at least 0.3 mg/dL, or urine output less than 0.5 mL/kg/hour for 6 to 12 hours
• Stage 2—an increase in serum creatinine of 2.0 to 2.9 times baseline or urine output less than 0.5 mmL/kg/hour for 12 or more hours
• Stage 3—an increase in serum creatinine of 3 times baseline, an absolute increase of at least 4 mg/dL, initiation of renal replacement therapy, urine output less than 0.3 mL/kg/hour for 24 or more hours, or anuria for 12 or more hours.¹⁵

Multiple factors associated with surgery may contribute to AKI, including hemodynamic instability, volume shifts, blood loss, use of heart-lung bypass, new medications, activation of the inflammatory cascade, oxidative stress, and anemia.

AKI in cardiac surgery

The incidence of AKI is high in cardiac surgery. In a meta-analysis of 46 studies (N = 242,000), its incidence in cardiopulmonary bypass surgery was about 18%, with 2.1% of patients needing renal replacement therapy.¹⁶ However, the incidence varied considerably from study to study, ranging from 1% to 53%, and was influenced by the definition of AKI, the type of cardiac surgery, and the patient population.¹⁶

Cardiac surgery-associated AKI adversely affects outcomes. Several studies have shown that cardiac surgery patients who develop AKI have higher rates of death and stroke.^16–21 More severe AKI confers higher mortality rates, with the highest mortality rate in patients who need renal replacement therapy, approximately 37%.¹⁷ Patients with cardiac surgery-associated AKI also have a longer hospital length of stay and significantly higher costs of care.^17,18

Long-term outcomes are also negatively affected by AKI. In cardiac surgery patients with AKI who had completely recovered renal function by the time they left the hospital, the 2-year incidence rate of CKD was 6.8%, significantly higher than the 0.2% rate in patients who did not develop AKI.¹⁹ The 2-year survival rates also were significantly worse for patients who developed postoperative AKI (82.3% vs 93.7%). Similarly, in patients undergoing CABG who had normal renal function before surgery, those who developed AKI postoperatively had significantly shorter long-term survival rates.²⁰ The effect does not require a large change in renal function. An increase in creatinine as small as 0.3 mg/dL has been associated with a higher rate of death and a long-term risk of end-stage renal disease that is 3 times higher.²¹

Cardiac surgery

CKD is a risk factor not only after cardiac surgery but also after percutaneous procedures. In a meta-analysis of 4,992 patients with CKD who underwent transcatheter aortic valve replacement, both moderate and severe CKD increased the odds of AKI, early stroke, the need for dialysis, and all-cause and cardiovascular mortality at 1 year.^22,23 Increased rates of AKI also have been found in patients with CKD undergoing CABG surgery.²⁴ These results point to a synergistic effect between AKI and CKD, with outcomes much worse in combination than alone.

In cardiac surgery, the most important patient risk factors associated with a higher incidence of postoperative AKI are age older than 75, CKD, preoperative heart failure, and prior myocardial infarction.^19,25 Diabetes is an additional independent risk factor, with type 1 conferring higher risk than type 2.²⁶ Preoperative use of angiotensin-converting enzyme (ACE) inhibitors may or may not be a risk factor for cardiac surgery-associated AKI, with some studies finding increased risk and others finding reduced rates.^27,28

Anemia, which may be related to either patient or surgical risk factors (eg, intraoperative blood loss), also increases the risk of AKI in cardiac surgery.^29,30 A retrospective study of CABG surgery patients found that intraoperative hemoglobin levels below 8 g/dL were associated with a 25% to 30% incidence of AKI, compared with 15% to 20% with hemoglobin levels above 9 g/dL.²⁹ Additionally, having severe hypotension (mean arterial pressure < 50 mm Hg) significantly increased the AKI rates in the low-hemoglobin group.²⁹ Similar results were reported in a later study.³⁰

Among surgical factors, several randomized controlled trials have shown that off-pump CABG is associated with a significantly lower risk of postoperative AKI than on-pump CABG; however, this difference did not translate into any long-term difference in mortality rates.^31,32 Longer cardiopulmonary bypass time is strongly associated with a higher incidence of AKI and postoperative death.³³

Noncardiac surgery

AKI is less common after noncardiac surgery; however, outcomes are severe in patients in whom it occurs. In a study of 15,102 noncardiac surgery patients, only 0.8% developed AKI and 0.1% required renal replacement therapy.³⁴

Risk factors after noncardiac surgery are similar to those after cardiac surgery (Table 3).^34–36 Factors with the greatest impact are older age, peripheral vascular occlusive disease, chronic obstructive pulmonary disease necessitating chronic bronchodilator therapy, high-risk surgery, hepatic disease, emergent or urgent surgery, and high body mass index.

Surgical risk factors include total vasopressor dose administered, use of a vasopressor infusion, and diuretic administration.³⁴ In addition, intraoperative hypotension is associated with a higher risk of AKI, major adverse cardiac events, and 30-day mortality.³⁷

Noncardiac surgery patients with postoperative AKI have significantly higher rates of 30-day readmissions, 1-year progression to end-stage renal disease, and mortality than patients who do not develop AKI.³⁵ Additionally, patients with AKI have significantly higher rates of cardiovascular complications (33.3% vs 11.3%) and death (6.1% vs 0.9%), as well as a significantly longer length of hospital stay.^34,36

CAN WE DECREASE THE IMPACT OF RENAL DISEASE IN SURGERY?

Before surgery, practitioners need to identify patients at risk of AKI, implement possible risk-reduction measures, and, afterward, treat it early in its course if it occurs.

The preoperative visit is the ideal time to assess a patient’s risk of postoperative renal dysfunction. Laboratory tests can identify risks based on surgery type, age, hypertension, the presence of CKD, and medications that affect renal function. However, the basic chemistry panel is abnormal in only 8.2% of patients and affects management in just 2.6%, requiring the clinician to target testing to patients at high risk.³⁸

Patients with a significant degree of renal dysfunction, particularly those previously undiagnosed, may benefit from additional preoperative testing and medication management. Perioperative management of medications that could adversely affect renal function should be carefully considered during the preoperative visit. In addition, the postoperative inpatient team needs to be informed about potentially nephrotoxic medications and medications that are renally cleared. Attention needs to be given to the renal impact of common perioperative medications such as nonsteroidal anti-inflammatory drugs, antibiotics, intravenous contrast, low-molecular-weight heparins, diuretics, ACE inhibitors, and angiotensin II receptor blockers. With the emphasis on opioid-sparing analgesics, it is particularly important to assess the risk of AKI if nonsteroidal anti-inflammatory drugs are part of the pain control plan.

Nephrology referral may help, especially for patients with a GFR less than 45 mL/min. This information enables more informed decision-making regarding the risks of adverse outcomes related to kidney disease.

WHAT TOOLS DO WE HAVE TO DIAGNOSE RENAL INJURY?

Several risk-prediction models have been developed to assess the postoperative risk of AKI in both cardiac and major noncardiac surgery patients. Although these models can identify risk factors, their clinical accuracy and utility have been questioned.

Biomarkers

Early diagnosis is the first step in managing AKI, allowing time to implement measures to minimize its impact.

Serum creatinine testing is widely used to measure renal function and diagnose AKI; however, it does not detect small reductions in renal function, and there is a time lag between renal insult and a rise in creatinine. The result is a delay to diagnosis of AKI.

Biomarkers other than creatinine have been studied for early detection of intraoperative and postoperative renal insult. These novel renal injury markers include the following:

Neutrophil gelatinase-associated lipocalin (NGAL). Two studies looked at plasma NGAL as an early marker of AKI in patients with CKD who were undergoing cardiac surgery.^39,40 One study found that by using NGAL instead of creatinine, postoperative AKI could be diagnosed an average of 20 hours earlier.³⁹ In addition, NGAL helped detect renal recovery earlier than creatinine.⁴⁰ The diagnostic cut-off values of NGAL were different for patients with CKD than for those without CKD.^39,40

Other novel markers include:

• Kidney injury marker 1
• N-acetyl-beta-D-glucosaminidase
• Cysteine C.

Although these biomarkers show some ability to detect renal injury, they provide only modest discrimination and are not widely available for clinical use.⁴¹ Current evidence does not support routine use of these markers in clinical settings.

Interventions to prevent or ameliorate the impact of CKD and AKI on surgical outcomes have been studied most extensively in cardiac surgery patients.

Aspirin. A retrospective study of 3,585 cardiac surgery patients with CKD found that preoperative aspirin use significantly lowered the incidence of postoperative AKI and 30-day mortality compared with patients not using aspirin.⁴² Aspirin use reduced 30-day mortality in CKD stages 1, 2, and 3 by 23.3%, 58%, and 70%, respectively. On the other hand, in the Perioperative Ischemic Evaluation (POISE) trial, in noncardiac surgery patients, neither aspirin nor clonidine started 2 to 4 hours preoperatively and continued up to 30 days after surgery altered the risk of AKI significantly more than placebo.⁴³

Statins have been ineffective in reducing the incidence of AKI in cardiac surgery patients. In fact, a meta-analysis of 8 interventional trials found an increased incidence of AKI in patients in whom statins were started perioperatively.⁴⁴ Erythropoietin was also found to be ineffective in the prevention of perioperative AKI in cardiac surgery patients in a separate study.⁴⁵

The evidence regarding other therapies has also varied.

N-acetylcysteine in high doses reduced the incidence of AKI in patients with CKD stage 3 and 4 undergoing CABG.⁴⁶ Another meta-analysis of 10 studies in cardiac surgery patients published recently did not show any benefit of N-acetylcysteine in reducing AKI.⁴⁷

Human atrial natriuretic peptide, given preoperatively to patients with CKD, reduced the acute and long-term creatinine rise as well as the number of cardiac events after CABG; however, it did not reduce mortality rates.⁴⁸

Renin-angiotensin system inhibitors, given preoperatively to patients with heart failure was associated with a decrease in the incidence of AKI in 1 study.⁴⁹

Dexmedetomidine is a highly selective alpha 2 adrenoreceptor agonist. A recent meta-analysis of 10 clinical trials found it beneficial in reducing the risk of perioperative AKI in cardiac surgery patients.⁵⁰ An earlier meta-analysis had similar results.⁵¹

Levosimendan is an inotropic vasodilator that improves cardiac output and renal perfusion in patients with systolic heart failure, and it has been hypothesized to decrease the risk of AKI after cardiac surgery. Previous data demonstrated that this drug reduced AKI and mortality; however, analysis was limited by small sample size and varying definitions of AKI.⁵² A recent meta-analysis showed that levosimendan was associated with a lower incidence of AKI but was also associated with an increased incidence of atrial fibrillation and no reduction in 30-day mortality.⁵³

Remote ischemic preconditioning is a procedure that subjects the kidneys to brief episodes of ischemia before surgery, protecting them when they are later subjected to prolonged ischemia or reperfusion injury. It has shown initial promising results in preventing AKI. In a randomized controlled trial in 240 patients at high risk of AKI, those who received remote ischemic preconditioning had an AKI incidence of 37.5% compared with 52.5% for controls (P = .02); however, the mortality rate was the same.⁵⁴ Similarly, remote ischemic preconditioning significantly lowered the incidence of AKI in nondiabetic patients undergoing CABG surgery compared with controls.⁵⁵

Fluid management. Renal perfusion is intimately related to the development of AKI, and there is evidence that both hypovolemia and excessive fluid resuscitation can increase the risk of AKI in noncardiac surgery patients.⁵⁶ Because of this, fluid management has also received attention in perioperative AKI. Goal-directed fluid management has been evaluated in noncardiac surgery patients, and it did not show any benefit in preventing AKI.⁵⁷ However, in a more recent retrospective study, postoperative positive fluid balance was associated with increased incidence of AKI compared with zero fluid balance. Negative fluid balance did not appear to have a detrimental effect.⁵⁸

RECOMMENDATIONS

No prophylactic therapy has yet been shown to definitively decrease the risk of postoperative AKI in all patients. Nevertheless, it is important to identify patients at risk during the preoperative visit, especially those with CKD. Many patients undergoing surgery have CKD, placing them at high risk of developing AKI in the perioperative period. The risk is particularly high with cardiac surgery.

Serum creatinine and urine output should be closely monitored perioperatively in at-risk patients. If AKI is diagnosed, practitioners need to identify and ameliorate the cause as early as possible.

Recommendations from KDIGO for perioperative prevention and management of AKI are listed in Table 4.¹⁵ These include avoiding additional nephrotoxic medications and adjusting the doses of renally cleared medications. Also, some patients may benefit from preoperative counseling and specialist referral.

Click here to access Federal Health Care Data Trends 2018

Table of Contents

• Veteran Demographics
• Vietnam Veterans
• Gulf War & Post 9/11 Veterans
• Cardiovascular Diseases
• Dermatology
• Neurology
• Oncology
• Respiratory Diseases
• Sleep Disorders
• References

Click here to access Federal Health Care Data Trends 2018

Table of Contents

• Veteran Demographics
• Vietnam Veterans
• Gulf War & Post 9/11 Veterans
• Cardiovascular Diseases
• Dermatology
• Neurology
• Oncology
• Respiratory Diseases
• Sleep Disorders
• References

Click here to access Federal Health Care Data Trends 2018

Table of Contents

• Veteran Demographics
• Vietnam Veterans
• Gulf War & Post 9/11 Veterans
• Cardiovascular Diseases
• Dermatology
• Neurology
• Oncology
• Respiratory Diseases
• Sleep Disorders
• References