A 46‐year‐old Mexican woman with acquired immune deficiency syndrome (AIDS), admitted for 6 months of diarrhea and failure to thrive, developed acute shortness of breath following colonoscopy. She reported dyspnea in the recumbent position, associated with a nonproductive cough, which improved with elevation of the head of the bed. She denied chest pain, palpitations, lightheadedness, hemoptysis, abdominal pain, nausea, and fever.

The approach to acute shortness of breath in hospitalized patients should include evaluation for life‐threatening cardiopulmonary processes. The patient should be assessed for cardiopulmonary process, including myocardial infarction, pulmonary embolism, aortic dissection, congestive heart failure, unstable arrhythmias, cardiac tamponade, and pneumothorax. The presence of orthopnea does suggest pulmonary congestion and a cardiac process. Given the timing of her symptoms, there is also concern for complications related to the colonoscopy, including aspiration pneumonitis, bronchospasm due to ethylene glycol, and methemoglobinemia from benzocaine used during the procedure.

The patient had been admitted the prior day for 6 months of diarrhea, weight loss, and failure to thrive. On admission, she was afebrile with a blood pressure of 110/50 mmHg and a pulse of 110 beats per minute; electrocardiogram (EKG) at the time revealed normal sinus rhythm. Her oxygen saturation was 100% on ambient air, and she had no complaints of cough, fevers, or dyspnea.

On admission, a peripherally inserted central catheter (PICC) was placed and total parenteral nutrition (TPN) was initiated. A gastroenterology consult was obtained, and colonoscopy was recommended to evaluate the cause of her chronic diarrhea. Overnight, the patient was started on polyethylene glycol electrolyte solution, with nothing else by mouth, and initiation of maintenance intravenous normal saline at 50 ml/hr in addition to her TPN. The patient expressed difficulty completing the colonoscopy preparation, but her preparation was acceptable to proceed with the procedure. She denied fever, chills, abdominal pain, and respiratory symptoms. She was taken down to endoscopy where she underwent conscious sedation, followed by an uneventful colonoscopy with mucosal biopsies. She subsequently was transported back to her hospital room in the supine position and almost immediately began to complain of mild shortness of breath. No aspiration event was witnessed following her procedure and transport.

Considering the patient's chronic diarrhea, there may be a unifying cause of both the gastrointestinal (GI) and pulmonary symptoms. Possibilities include infectious causes (Toxoplasma gondii and Trypanosoma cruzi), infiltrative diseases (amyloidosis), and metabolic processes (hyperthyroidism). More specifically, T. cruzi can cause dilated cardiomyopathy, with subsequent congestive heart failure and associated pulmonary symptoms; furthermore, it can lead to a dilated colon with abnormal bowel movements. Opportunistic infections, including Microsporidia, Cryptosporidium, Mycobacterium avium complex (MAC), and cytomegalovirus (CMV) should be considered. MAC and CMV can present with non‐bloody diarrhea and evolve into respiratory illnesses. Lastly, human immunodeficiency virus (HIV) is known to involve multiple organ systems, including the heart and gastrointestinal tract. History of prior cardiac or pulmonary disease, CD4 count and viral load, use of antiretroviral and prophylactic medications, and recent travel should be obtained.

Thirteen years previously, the patient was diagnosed with HIV, and subsequently developed AIDS with thrush and uncomplicated CMV viremia. At that time, highly active antiretroviral therapy (HAART)was initiated, but she was intolerant of her medications and received therapy intermittently. Her past medical history included multiple fractures secondary to osteoporosis. She denied any history of respiratory or cardiac symptoms. The patient was born in rural Mexico and immigrated to the United States 20 years prior. Her last visit to Mexico occurred 2 months prior to admission, and 4 months following the development of chronic diarrhea. Previously, she worked as a housekeeper and was not aware of any toxic exposures during cleaning. She denied a history of alcohol or recreational drug use. Despite her generalized weakness, her baseline functional status included performing all activities of daily living without symptoms.

Over the prior 6 months, the patient had developed diffuse watery diarrhea, associated with a 20‐pound weight loss. Stool evaluation, 1 week prior, was negative for Clostridium difficile, Microsporidia, Isospora, Cryptosporidium, Escherichia coli, Campylobacter, and ova and parasites. Her CD4 count was 8 cells per cubic millimeter.

The low CD4 count predisposes the patient to all opportunistic infections. Considering the history of CMV viremia, there is likelihood of reactivation with viremia and colitis, leading to chronic diarrhea and pneumonia. Disseminated MAC infection is also a consideration and would account for wasting, diarrhea, and dyspnea. However, it is important to note that the acute onset of dyspnea is atypical for CMV and MAC infections.

On physical exam, she was a thin woman with temporal wasting, in mild respiratory distress. Her temperature was 37.3C, blood pressure 133/55 mmHg, heart rate 140 beats/min, respiratory rate 22 breaths/min, and oxygen saturation 89% on room air. Her oropharynx was clear and without acral cyanosis. Use of accessory muscles for breathing was noted. The trachea was midline and no lymphadenopathy or thyromegaly were present. Her jugular venous pulse was normal. Cardiac exam revealed tachycardia with a new S4 gallop. A prominent apical impulse was noted. No murmurs or rubs were appreciated. There was no pulsus paradoxus. Her radial, femoral, and dorsalis pedis pulses were 2+ without delay. Her lung exam revealed inspiratory crackles involving the lower one‐third of both lungs. The lower extremities revealed 2+ pitting edema to the knees. The rest of her exam, including her neurologic evaluation, was unremarkable.

These clinical findings are consistent with left‐sided heart failure, concerning for ischemic injury or structural disorders of the heart. It is possible that the patient has had progressive heart failure, which is now unmasked by the volume received with TPN and endoscopy. If the heart failure has been longstanding, one has to consider potential non‐ischemic causes of cardiomyopathy, including infectious etiologies such as HIV, Epstein‐Barr virus (EBV), coxsackie virus, CMV, Toxoplasma gondii, and Trypanosoma cruzi; alcohol‐associated; and pericardial disease with Mycobacterium tuberculosis (MTB). Toxoplasma should be evaluated if the patient has exposure to cats. Considering her country of origin and travel history, risk factors for trypanosomiasis and MTB should be assessed.

At this point, the patient's respiratory failure should be aggressively addressed. Supplemental oxygen should be administered. She should be evaluated for acute coronary syndrome with an EKG and serial cardiac enzymes. A chest x‐ray should be obtained to grossly evaluate for pulmonary, pericardial, and aortic illnesses. Brain natriuretic peptide (BNP) levels should also be sent. Considering the evidence of volume overload and her HIV status, liver function tests, serum electrolytes, and urinalysis should be sent to exclude liver and renal involvement.

The patient was placed on 2 liters of oxygen by nasal cannula with resolution of her symptoms and improvement in her oxygen saturation to 95%. An EKG demonstrated sinus tachycardia, without evidence of ischemia. Metabolic panel revealed sodium 134 mmol/L; potassium 4.3 mmol/L; chloride 105 mmol/L; bicarbonate 16 mmol/L; creatinine 0.6 mg/dL, and liver function tests were within normal limits. Her troponin level was within the normal range for a negative value, and BNP was 823 pg/ml (normal 100). The complete blood count demonstrated leukopenia and anemia (hemoglobin 9.8 g/dL), which were unchanged from admission. Urinalysis was negative. A portable chest x‐ray demonstrated vascular congestion and mild pulmonary edema, without evidence of pneumothorax or pleural effusion.

The significantly elevated BNP and pulmonary vascular congestion seen on chest x‐ray confirm the clinical diagnosis of heart failure. However, the negative troponin and unremarkable EKG suggest a non‐ischemic cause for her symptoms. An echocardiogram should be obtained with specific emphasis on the presence of valve regurgitation, pericardial effusion, and ventricular/atrial thickening consistent with infiltrative disorders. Thyroid stimulating hormone (TSH) and serologies for infectious agents, including, T. cruzi, HIV, CMV, and Toxoplasmosis gondii, should also be sent. The patient should receive intravenous loop diuretics to improve her cardiac dynamics and pulmonary edema.

Intravenous furosemide was administered. Her symptoms improved and oxygen saturation on room air was 92%. An echocardiogram revealed global hypokinesis with left ventricular ejection fraction (LVEF) of 35% to 40%. There was no evidence of an underlying valvular or infiltrative process. TSH was normal. T. cruzi antibodies were sent.

The echocardiogram did not reveal an underlying structural heart abnormality. Infiltrative cardiomyopathies do not typically demonstrate global hypokinesis on echocardiogram, particularly without evidence of ventricular wall thickening or increased echogenicity, that can be seen in amyloid and sarcoid cardiomyopathies. Therefore, infiltrative cardiomyopathy is unlikely to be a cause of this patient's heart failure. The rapid improvement of her symptoms with furosemide decreases the likelihood of infectious causes for her acute decompensation. In reviewing the patient's history, she had developed severe chronic diarrhea associated with poor oral intake and a 20‐pound weight loss prior to hospitalization. These symptoms, along with a history of osteoporosis at an early age without traditional risk factors, indicate a state of severe malnutrition, placing her at risk for thiamine deficiency. Checking the thiamine level would be appropriate.

Considering the patient's long history of malnutrition and negative infectious and ischemic evaluation, she was empirically treated for wet beriberi with thiamine supplementation through her TPN. A serum thiamine B₁ was obtained prior to supplementation. A vitamin D 25OH level was also sent, which was 15 ng/mL (normal >30 ng/mL), further suggesting malnutrition.

The patient continued to improve and furosemide was discontinued. Her initial serum thiamine level was 49 nmol/L (reference range: 70‐180 nmol/L). A repeat echocardiogram 5 days later revealed resolution of her systolic dysfunction and regional wall motion abnormality. The LVEF improved to 60%. Her colonoscopy biopsies revealed evidence of HIV enteropathy and CMV inclusion bodies. Her CMV viral load was 1223 genomes/mL. The T. cruzi antibodies were negative. She was restarted on HAART and ganciclovir. She continued to have diarrhea and was discharged home with TPN. Her serum thiamine level at discharge was 123 nmol/L.

Heart failure due to thiamine deficiency, or wet beriberi, was diagnosed considering the rapid clinical improvement in cardiac function after initiating thiamine therapy. While HIV cardiomyopathy could have contributed to heart failure in this patient, it is unlikely to improve so significantly over such a brief period of time.

DISCUSSION

Beriberi is a disease caused by severe thiamine deficiency. In fact, thiamine, also known as vitamin B₁, was first named the anti‐beriberi factor in 1926. However, the earliest descriptions of beriberi can be found in Chinese medical texts dating back to 2697 _BC.1 Beriberi is most commonly seen in Asia, where the diet is high in polished rice and the thiamine‐containing rice germs and husks have been removed. In the United States, thiamine‐enriched bread has virtually abolished the disease, except in severely malnourished populations such as alcoholics, those on fad diets, and patients with chronic diarrhea. Beriberi may also occur in patients with altered intestinal absorption such as post‐bariatric surgery patients.2 In 1985, the first case of beriberi as a complication of TPN without vitamin supplementation was reported.3 Subsequent cases of Wernicke's encephalopathy and beriberi have been noted in patients with gastrointestinal diseases and malabsorption on chronic TPN. More recently, thiamine deficiency has also been recognized in patients on long‐term diuretic therapy, as diuretics increase urinary excretion of this water‐soluble vitamin.4, 5 Since there is limited tissue storage of thiamine and its biologic half‐life is 10 to 20 days, high‐risk patients can develop thiamine deficiency within 4 weeks of initiation of diuretic therapy.6

Beriberi is classically divided into 2 types: wet, characterized by congestive heart failure, and dry, manifested as a symmetric peripheral neuropathy with both sensory and motor impairments.7 These 2 types of beriberi can coexist in the same patient; however, it is unclear why both types occur in some patients and not in others. Wet beriberi, also known as beriberi cardiomyopathy, typically presents as high‐output heart failure secondary to vasodilation, with a compensatory increase in blood volume and tachycardia.8 This state eventually leads to myocardial injury with systolic dysfunction and development of a low‐output state.8 Patients experience hypotension, lactic acidosis, and eventually fulminant vascular collapse. Although minor EKG changes such as sinus tachycardia, low‐voltage ventricular complexes, QT prolongation, and biphasic or inverted T waves are not uncommon in beriberi cardiomyopathy, major EKG changes, such as ST segment elevations and tall or deeply inverted T waves, are rare. Similarly, troponin elevation in beriberi cardiomyopathy is uncommon, but has been described.6

The pathogenesis of heart failure in beriberi is multifactorial. Thiamine is required for glucose to enter the Krebs cycle for aerobic metabolism, serving as a catalyst in the conversion of pyruvate to acetyl‐CoA. Without thiamine, anaerobic metabolism occurs, leading to the development of lactic acidosis and cellular malfunction. In fact, severe metabolic acidosis with serum pH values as low as 6.70 have been reported in cases of fulminant beriberi (although it is unclear if the lactic acidosis is mostly from anaerobic metabolism or from the low‐output state ultimately caused by thiamine deficiency).3

Laboratory diagnosis of thiamine deficiency, based on measurements of thiamine stores and metabolites, is often fraught with error and therefore unreliable. Serum pyruvate and lactate levels are commonly measured, and while elevated levels may be sensitive for thiamine deficiency, they are nonspecific. Measurement of whole blood thiamine is easy and the test is widely available; however, a low blood thiamine concentration is not always a sensitive indicator of deficiency since less than 1% of total body thiamine is found in whole blood.9 Additionally, this value may also be artificially elevated by thiamine intake immediately preceding the measurement. Urinary thiamine excretion has been proposed as a more accurate measurement, but this laboratory test is also problematic since urinary thiamine excretion reflects dietary intake more than total body stores.9 Erythrocyte transketolase activity (ETKA) is a functional enzyme test in which transketolase uses thiamine pyrophosphate as a catalyzer. This may be a more reliable measurement since red blood cells are among the first cells to be affected by thiamine depletion.9 Although a low ETKA level often indicates thiamine deficiency, this test is influenced by the hemoglobin concentration, and it is not widely available. Thus, the diagnosis of wet beriberi is usually made on the basis of rapid response to thiamine replacement.

Similar to the patient discussed, the clinical improvement in wet beriberi occurs within hours of treatment. There is an initial elevation in blood pressure and resolution of acidosis, followed by decrease in heart rate and normalization of cardiac output. Overall cardiac function improves within 24 to 48 hours after treatment, and return to a normal hemodynamic condition often occurs within 2 weeks of the start of treatment.10

There are no well‐established guidelines for the treatment of patients with beriberi, but general recommendations are an initial loading dose of intravenous thiamine 100 to 500 mg followed by 25 to 100 mg orally for 7 to 14 days.1 Thereafter, the daily thiamine requirement can be calculated based upon total caloric intake. The current recommendations in the United States are 0.5 mg of thiamine per 1000 kcal.1 However, one must consider whether a patient has impaired intestinal absorption or increased urinary losses when determining an appropriate maintenance dose.

Chronic malnutrition can lead to significant morbidity and mortality. Prior to admission, this patient already exhibited signs of severe malnutrition with a history of multiple pathologic fractures and diagnosis of osteoporosis. Considering her age and lack of risk factors for bone disease, osteoporosis suggests vitamin D deficiency. In this patient with chronic diarrhea caused by CMV, it is unlikely that a selective absorptive deficiency would occur. When the common causes of acute heart failure following volume challenge were excluded, the diagnosis of thiamine deficiency became more likely. Fortunately, an empiric trial of intravenous thiamine resulted in diagnosis by treatment and improvement of her cardiac function.

A 46‐year‐old Mexican woman with acquired immune deficiency syndrome (AIDS), admitted for 6 months of diarrhea and failure to thrive, developed acute shortness of breath following colonoscopy. She reported dyspnea in the recumbent position, associated with a nonproductive cough, which improved with elevation of the head of the bed. She denied chest pain, palpitations, lightheadedness, hemoptysis, abdominal pain, nausea, and fever.

The approach to acute shortness of breath in hospitalized patients should include evaluation for life‐threatening cardiopulmonary processes. The patient should be assessed for cardiopulmonary process, including myocardial infarction, pulmonary embolism, aortic dissection, congestive heart failure, unstable arrhythmias, cardiac tamponade, and pneumothorax. The presence of orthopnea does suggest pulmonary congestion and a cardiac process. Given the timing of her symptoms, there is also concern for complications related to the colonoscopy, including aspiration pneumonitis, bronchospasm due to ethylene glycol, and methemoglobinemia from benzocaine used during the procedure.

The patient had been admitted the prior day for 6 months of diarrhea, weight loss, and failure to thrive. On admission, she was afebrile with a blood pressure of 110/50 mmHg and a pulse of 110 beats per minute; electrocardiogram (EKG) at the time revealed normal sinus rhythm. Her oxygen saturation was 100% on ambient air, and she had no complaints of cough, fevers, or dyspnea.

On admission, a peripherally inserted central catheter (PICC) was placed and total parenteral nutrition (TPN) was initiated. A gastroenterology consult was obtained, and colonoscopy was recommended to evaluate the cause of her chronic diarrhea. Overnight, the patient was started on polyethylene glycol electrolyte solution, with nothing else by mouth, and initiation of maintenance intravenous normal saline at 50 ml/hr in addition to her TPN. The patient expressed difficulty completing the colonoscopy preparation, but her preparation was acceptable to proceed with the procedure. She denied fever, chills, abdominal pain, and respiratory symptoms. She was taken down to endoscopy where she underwent conscious sedation, followed by an uneventful colonoscopy with mucosal biopsies. She subsequently was transported back to her hospital room in the supine position and almost immediately began to complain of mild shortness of breath. No aspiration event was witnessed following her procedure and transport.

Considering the patient's chronic diarrhea, there may be a unifying cause of both the gastrointestinal (GI) and pulmonary symptoms. Possibilities include infectious causes (Toxoplasma gondii and Trypanosoma cruzi), infiltrative diseases (amyloidosis), and metabolic processes (hyperthyroidism). More specifically, T. cruzi can cause dilated cardiomyopathy, with subsequent congestive heart failure and associated pulmonary symptoms; furthermore, it can lead to a dilated colon with abnormal bowel movements. Opportunistic infections, including Microsporidia, Cryptosporidium, Mycobacterium avium complex (MAC), and cytomegalovirus (CMV) should be considered. MAC and CMV can present with non‐bloody diarrhea and evolve into respiratory illnesses. Lastly, human immunodeficiency virus (HIV) is known to involve multiple organ systems, including the heart and gastrointestinal tract. History of prior cardiac or pulmonary disease, CD4 count and viral load, use of antiretroviral and prophylactic medications, and recent travel should be obtained.

Thirteen years previously, the patient was diagnosed with HIV, and subsequently developed AIDS with thrush and uncomplicated CMV viremia. At that time, highly active antiretroviral therapy (HAART)was initiated, but she was intolerant of her medications and received therapy intermittently. Her past medical history included multiple fractures secondary to osteoporosis. She denied any history of respiratory or cardiac symptoms. The patient was born in rural Mexico and immigrated to the United States 20 years prior. Her last visit to Mexico occurred 2 months prior to admission, and 4 months following the development of chronic diarrhea. Previously, she worked as a housekeeper and was not aware of any toxic exposures during cleaning. She denied a history of alcohol or recreational drug use. Despite her generalized weakness, her baseline functional status included performing all activities of daily living without symptoms.

Over the prior 6 months, the patient had developed diffuse watery diarrhea, associated with a 20‐pound weight loss. Stool evaluation, 1 week prior, was negative for Clostridium difficile, Microsporidia, Isospora, Cryptosporidium, Escherichia coli, Campylobacter, and ova and parasites. Her CD4 count was 8 cells per cubic millimeter.

The low CD4 count predisposes the patient to all opportunistic infections. Considering the history of CMV viremia, there is likelihood of reactivation with viremia and colitis, leading to chronic diarrhea and pneumonia. Disseminated MAC infection is also a consideration and would account for wasting, diarrhea, and dyspnea. However, it is important to note that the acute onset of dyspnea is atypical for CMV and MAC infections.

On physical exam, she was a thin woman with temporal wasting, in mild respiratory distress. Her temperature was 37.3C, blood pressure 133/55 mmHg, heart rate 140 beats/min, respiratory rate 22 breaths/min, and oxygen saturation 89% on room air. Her oropharynx was clear and without acral cyanosis. Use of accessory muscles for breathing was noted. The trachea was midline and no lymphadenopathy or thyromegaly were present. Her jugular venous pulse was normal. Cardiac exam revealed tachycardia with a new S4 gallop. A prominent apical impulse was noted. No murmurs or rubs were appreciated. There was no pulsus paradoxus. Her radial, femoral, and dorsalis pedis pulses were 2+ without delay. Her lung exam revealed inspiratory crackles involving the lower one‐third of both lungs. The lower extremities revealed 2+ pitting edema to the knees. The rest of her exam, including her neurologic evaluation, was unremarkable.

These clinical findings are consistent with left‐sided heart failure, concerning for ischemic injury or structural disorders of the heart. It is possible that the patient has had progressive heart failure, which is now unmasked by the volume received with TPN and endoscopy. If the heart failure has been longstanding, one has to consider potential non‐ischemic causes of cardiomyopathy, including infectious etiologies such as HIV, Epstein‐Barr virus (EBV), coxsackie virus, CMV, Toxoplasma gondii, and Trypanosoma cruzi; alcohol‐associated; and pericardial disease with Mycobacterium tuberculosis (MTB). Toxoplasma should be evaluated if the patient has exposure to cats. Considering her country of origin and travel history, risk factors for trypanosomiasis and MTB should be assessed.

At this point, the patient's respiratory failure should be aggressively addressed. Supplemental oxygen should be administered. She should be evaluated for acute coronary syndrome with an EKG and serial cardiac enzymes. A chest x‐ray should be obtained to grossly evaluate for pulmonary, pericardial, and aortic illnesses. Brain natriuretic peptide (BNP) levels should also be sent. Considering the evidence of volume overload and her HIV status, liver function tests, serum electrolytes, and urinalysis should be sent to exclude liver and renal involvement.

The patient was placed on 2 liters of oxygen by nasal cannula with resolution of her symptoms and improvement in her oxygen saturation to 95%. An EKG demonstrated sinus tachycardia, without evidence of ischemia. Metabolic panel revealed sodium 134 mmol/L; potassium 4.3 mmol/L; chloride 105 mmol/L; bicarbonate 16 mmol/L; creatinine 0.6 mg/dL, and liver function tests were within normal limits. Her troponin level was within the normal range for a negative value, and BNP was 823 pg/ml (normal 100). The complete blood count demonstrated leukopenia and anemia (hemoglobin 9.8 g/dL), which were unchanged from admission. Urinalysis was negative. A portable chest x‐ray demonstrated vascular congestion and mild pulmonary edema, without evidence of pneumothorax or pleural effusion.

The significantly elevated BNP and pulmonary vascular congestion seen on chest x‐ray confirm the clinical diagnosis of heart failure. However, the negative troponin and unremarkable EKG suggest a non‐ischemic cause for her symptoms. An echocardiogram should be obtained with specific emphasis on the presence of valve regurgitation, pericardial effusion, and ventricular/atrial thickening consistent with infiltrative disorders. Thyroid stimulating hormone (TSH) and serologies for infectious agents, including, T. cruzi, HIV, CMV, and Toxoplasmosis gondii, should also be sent. The patient should receive intravenous loop diuretics to improve her cardiac dynamics and pulmonary edema.

Intravenous furosemide was administered. Her symptoms improved and oxygen saturation on room air was 92%. An echocardiogram revealed global hypokinesis with left ventricular ejection fraction (LVEF) of 35% to 40%. There was no evidence of an underlying valvular or infiltrative process. TSH was normal. T. cruzi antibodies were sent.

The echocardiogram did not reveal an underlying structural heart abnormality. Infiltrative cardiomyopathies do not typically demonstrate global hypokinesis on echocardiogram, particularly without evidence of ventricular wall thickening or increased echogenicity, that can be seen in amyloid and sarcoid cardiomyopathies. Therefore, infiltrative cardiomyopathy is unlikely to be a cause of this patient's heart failure. The rapid improvement of her symptoms with furosemide decreases the likelihood of infectious causes for her acute decompensation. In reviewing the patient's history, she had developed severe chronic diarrhea associated with poor oral intake and a 20‐pound weight loss prior to hospitalization. These symptoms, along with a history of osteoporosis at an early age without traditional risk factors, indicate a state of severe malnutrition, placing her at risk for thiamine deficiency. Checking the thiamine level would be appropriate.

Considering the patient's long history of malnutrition and negative infectious and ischemic evaluation, she was empirically treated for wet beriberi with thiamine supplementation through her TPN. A serum thiamine B₁ was obtained prior to supplementation. A vitamin D 25OH level was also sent, which was 15 ng/mL (normal >30 ng/mL), further suggesting malnutrition.

The patient continued to improve and furosemide was discontinued. Her initial serum thiamine level was 49 nmol/L (reference range: 70‐180 nmol/L). A repeat echocardiogram 5 days later revealed resolution of her systolic dysfunction and regional wall motion abnormality. The LVEF improved to 60%. Her colonoscopy biopsies revealed evidence of HIV enteropathy and CMV inclusion bodies. Her CMV viral load was 1223 genomes/mL. The T. cruzi antibodies were negative. She was restarted on HAART and ganciclovir. She continued to have diarrhea and was discharged home with TPN. Her serum thiamine level at discharge was 123 nmol/L.

Heart failure due to thiamine deficiency, or wet beriberi, was diagnosed considering the rapid clinical improvement in cardiac function after initiating thiamine therapy. While HIV cardiomyopathy could have contributed to heart failure in this patient, it is unlikely to improve so significantly over such a brief period of time.

DISCUSSION

Beriberi is a disease caused by severe thiamine deficiency. In fact, thiamine, also known as vitamin B₁, was first named the anti‐beriberi factor in 1926. However, the earliest descriptions of beriberi can be found in Chinese medical texts dating back to 2697 _BC.1 Beriberi is most commonly seen in Asia, where the diet is high in polished rice and the thiamine‐containing rice germs and husks have been removed. In the United States, thiamine‐enriched bread has virtually abolished the disease, except in severely malnourished populations such as alcoholics, those on fad diets, and patients with chronic diarrhea. Beriberi may also occur in patients with altered intestinal absorption such as post‐bariatric surgery patients.2 In 1985, the first case of beriberi as a complication of TPN without vitamin supplementation was reported.3 Subsequent cases of Wernicke's encephalopathy and beriberi have been noted in patients with gastrointestinal diseases and malabsorption on chronic TPN. More recently, thiamine deficiency has also been recognized in patients on long‐term diuretic therapy, as diuretics increase urinary excretion of this water‐soluble vitamin.4, 5 Since there is limited tissue storage of thiamine and its biologic half‐life is 10 to 20 days, high‐risk patients can develop thiamine deficiency within 4 weeks of initiation of diuretic therapy.6

Beriberi is classically divided into 2 types: wet, characterized by congestive heart failure, and dry, manifested as a symmetric peripheral neuropathy with both sensory and motor impairments.7 These 2 types of beriberi can coexist in the same patient; however, it is unclear why both types occur in some patients and not in others. Wet beriberi, also known as beriberi cardiomyopathy, typically presents as high‐output heart failure secondary to vasodilation, with a compensatory increase in blood volume and tachycardia.8 This state eventually leads to myocardial injury with systolic dysfunction and development of a low‐output state.8 Patients experience hypotension, lactic acidosis, and eventually fulminant vascular collapse. Although minor EKG changes such as sinus tachycardia, low‐voltage ventricular complexes, QT prolongation, and biphasic or inverted T waves are not uncommon in beriberi cardiomyopathy, major EKG changes, such as ST segment elevations and tall or deeply inverted T waves, are rare. Similarly, troponin elevation in beriberi cardiomyopathy is uncommon, but has been described.6

The pathogenesis of heart failure in beriberi is multifactorial. Thiamine is required for glucose to enter the Krebs cycle for aerobic metabolism, serving as a catalyst in the conversion of pyruvate to acetyl‐CoA. Without thiamine, anaerobic metabolism occurs, leading to the development of lactic acidosis and cellular malfunction. In fact, severe metabolic acidosis with serum pH values as low as 6.70 have been reported in cases of fulminant beriberi (although it is unclear if the lactic acidosis is mostly from anaerobic metabolism or from the low‐output state ultimately caused by thiamine deficiency).3

Laboratory diagnosis of thiamine deficiency, based on measurements of thiamine stores and metabolites, is often fraught with error and therefore unreliable. Serum pyruvate and lactate levels are commonly measured, and while elevated levels may be sensitive for thiamine deficiency, they are nonspecific. Measurement of whole blood thiamine is easy and the test is widely available; however, a low blood thiamine concentration is not always a sensitive indicator of deficiency since less than 1% of total body thiamine is found in whole blood.9 Additionally, this value may also be artificially elevated by thiamine intake immediately preceding the measurement. Urinary thiamine excretion has been proposed as a more accurate measurement, but this laboratory test is also problematic since urinary thiamine excretion reflects dietary intake more than total body stores.9 Erythrocyte transketolase activity (ETKA) is a functional enzyme test in which transketolase uses thiamine pyrophosphate as a catalyzer. This may be a more reliable measurement since red blood cells are among the first cells to be affected by thiamine depletion.9 Although a low ETKA level often indicates thiamine deficiency, this test is influenced by the hemoglobin concentration, and it is not widely available. Thus, the diagnosis of wet beriberi is usually made on the basis of rapid response to thiamine replacement.

Similar to the patient discussed, the clinical improvement in wet beriberi occurs within hours of treatment. There is an initial elevation in blood pressure and resolution of acidosis, followed by decrease in heart rate and normalization of cardiac output. Overall cardiac function improves within 24 to 48 hours after treatment, and return to a normal hemodynamic condition often occurs within 2 weeks of the start of treatment.10

There are no well‐established guidelines for the treatment of patients with beriberi, but general recommendations are an initial loading dose of intravenous thiamine 100 to 500 mg followed by 25 to 100 mg orally for 7 to 14 days.1 Thereafter, the daily thiamine requirement can be calculated based upon total caloric intake. The current recommendations in the United States are 0.5 mg of thiamine per 1000 kcal.1 However, one must consider whether a patient has impaired intestinal absorption or increased urinary losses when determining an appropriate maintenance dose.

Chronic malnutrition can lead to significant morbidity and mortality. Prior to admission, this patient already exhibited signs of severe malnutrition with a history of multiple pathologic fractures and diagnosis of osteoporosis. Considering her age and lack of risk factors for bone disease, osteoporosis suggests vitamin D deficiency. In this patient with chronic diarrhea caused by CMV, it is unlikely that a selective absorptive deficiency would occur. When the common causes of acute heart failure following volume challenge were excluded, the diagnosis of thiamine deficiency became more likely. Fortunately, an empiric trial of intravenous thiamine resulted in diagnosis by treatment and improvement of her cardiac function.

A 46‐year‐old Mexican woman with acquired immune deficiency syndrome (AIDS), admitted for 6 months of diarrhea and failure to thrive, developed acute shortness of breath following colonoscopy. She reported dyspnea in the recumbent position, associated with a nonproductive cough, which improved with elevation of the head of the bed. She denied chest pain, palpitations, lightheadedness, hemoptysis, abdominal pain, nausea, and fever.

The approach to acute shortness of breath in hospitalized patients should include evaluation for life‐threatening cardiopulmonary processes. The patient should be assessed for cardiopulmonary process, including myocardial infarction, pulmonary embolism, aortic dissection, congestive heart failure, unstable arrhythmias, cardiac tamponade, and pneumothorax. The presence of orthopnea does suggest pulmonary congestion and a cardiac process. Given the timing of her symptoms, there is also concern for complications related to the colonoscopy, including aspiration pneumonitis, bronchospasm due to ethylene glycol, and methemoglobinemia from benzocaine used during the procedure.

The patient had been admitted the prior day for 6 months of diarrhea, weight loss, and failure to thrive. On admission, she was afebrile with a blood pressure of 110/50 mmHg and a pulse of 110 beats per minute; electrocardiogram (EKG) at the time revealed normal sinus rhythm. Her oxygen saturation was 100% on ambient air, and she had no complaints of cough, fevers, or dyspnea.

On admission, a peripherally inserted central catheter (PICC) was placed and total parenteral nutrition (TPN) was initiated. A gastroenterology consult was obtained, and colonoscopy was recommended to evaluate the cause of her chronic diarrhea. Overnight, the patient was started on polyethylene glycol electrolyte solution, with nothing else by mouth, and initiation of maintenance intravenous normal saline at 50 ml/hr in addition to her TPN. The patient expressed difficulty completing the colonoscopy preparation, but her preparation was acceptable to proceed with the procedure. She denied fever, chills, abdominal pain, and respiratory symptoms. She was taken down to endoscopy where she underwent conscious sedation, followed by an uneventful colonoscopy with mucosal biopsies. She subsequently was transported back to her hospital room in the supine position and almost immediately began to complain of mild shortness of breath. No aspiration event was witnessed following her procedure and transport.

Considering the patient's chronic diarrhea, there may be a unifying cause of both the gastrointestinal (GI) and pulmonary symptoms. Possibilities include infectious causes (Toxoplasma gondii and Trypanosoma cruzi), infiltrative diseases (amyloidosis), and metabolic processes (hyperthyroidism). More specifically, T. cruzi can cause dilated cardiomyopathy, with subsequent congestive heart failure and associated pulmonary symptoms; furthermore, it can lead to a dilated colon with abnormal bowel movements. Opportunistic infections, including Microsporidia, Cryptosporidium, Mycobacterium avium complex (MAC), and cytomegalovirus (CMV) should be considered. MAC and CMV can present with non‐bloody diarrhea and evolve into respiratory illnesses. Lastly, human immunodeficiency virus (HIV) is known to involve multiple organ systems, including the heart and gastrointestinal tract. History of prior cardiac or pulmonary disease, CD4 count and viral load, use of antiretroviral and prophylactic medications, and recent travel should be obtained.

Thirteen years previously, the patient was diagnosed with HIV, and subsequently developed AIDS with thrush and uncomplicated CMV viremia. At that time, highly active antiretroviral therapy (HAART)was initiated, but she was intolerant of her medications and received therapy intermittently. Her past medical history included multiple fractures secondary to osteoporosis. She denied any history of respiratory or cardiac symptoms. The patient was born in rural Mexico and immigrated to the United States 20 years prior. Her last visit to Mexico occurred 2 months prior to admission, and 4 months following the development of chronic diarrhea. Previously, she worked as a housekeeper and was not aware of any toxic exposures during cleaning. She denied a history of alcohol or recreational drug use. Despite her generalized weakness, her baseline functional status included performing all activities of daily living without symptoms.

Over the prior 6 months, the patient had developed diffuse watery diarrhea, associated with a 20‐pound weight loss. Stool evaluation, 1 week prior, was negative for Clostridium difficile, Microsporidia, Isospora, Cryptosporidium, Escherichia coli, Campylobacter, and ova and parasites. Her CD4 count was 8 cells per cubic millimeter.

The low CD4 count predisposes the patient to all opportunistic infections. Considering the history of CMV viremia, there is likelihood of reactivation with viremia and colitis, leading to chronic diarrhea and pneumonia. Disseminated MAC infection is also a consideration and would account for wasting, diarrhea, and dyspnea. However, it is important to note that the acute onset of dyspnea is atypical for CMV and MAC infections.

On physical exam, she was a thin woman with temporal wasting, in mild respiratory distress. Her temperature was 37.3C, blood pressure 133/55 mmHg, heart rate 140 beats/min, respiratory rate 22 breaths/min, and oxygen saturation 89% on room air. Her oropharynx was clear and without acral cyanosis. Use of accessory muscles for breathing was noted. The trachea was midline and no lymphadenopathy or thyromegaly were present. Her jugular venous pulse was normal. Cardiac exam revealed tachycardia with a new S4 gallop. A prominent apical impulse was noted. No murmurs or rubs were appreciated. There was no pulsus paradoxus. Her radial, femoral, and dorsalis pedis pulses were 2+ without delay. Her lung exam revealed inspiratory crackles involving the lower one‐third of both lungs. The lower extremities revealed 2+ pitting edema to the knees. The rest of her exam, including her neurologic evaluation, was unremarkable.

These clinical findings are consistent with left‐sided heart failure, concerning for ischemic injury or structural disorders of the heart. It is possible that the patient has had progressive heart failure, which is now unmasked by the volume received with TPN and endoscopy. If the heart failure has been longstanding, one has to consider potential non‐ischemic causes of cardiomyopathy, including infectious etiologies such as HIV, Epstein‐Barr virus (EBV), coxsackie virus, CMV, Toxoplasma gondii, and Trypanosoma cruzi; alcohol‐associated; and pericardial disease with Mycobacterium tuberculosis (MTB). Toxoplasma should be evaluated if the patient has exposure to cats. Considering her country of origin and travel history, risk factors for trypanosomiasis and MTB should be assessed.

At this point, the patient's respiratory failure should be aggressively addressed. Supplemental oxygen should be administered. She should be evaluated for acute coronary syndrome with an EKG and serial cardiac enzymes. A chest x‐ray should be obtained to grossly evaluate for pulmonary, pericardial, and aortic illnesses. Brain natriuretic peptide (BNP) levels should also be sent. Considering the evidence of volume overload and her HIV status, liver function tests, serum electrolytes, and urinalysis should be sent to exclude liver and renal involvement.

The patient was placed on 2 liters of oxygen by nasal cannula with resolution of her symptoms and improvement in her oxygen saturation to 95%. An EKG demonstrated sinus tachycardia, without evidence of ischemia. Metabolic panel revealed sodium 134 mmol/L; potassium 4.3 mmol/L; chloride 105 mmol/L; bicarbonate 16 mmol/L; creatinine 0.6 mg/dL, and liver function tests were within normal limits. Her troponin level was within the normal range for a negative value, and BNP was 823 pg/ml (normal 100). The complete blood count demonstrated leukopenia and anemia (hemoglobin 9.8 g/dL), which were unchanged from admission. Urinalysis was negative. A portable chest x‐ray demonstrated vascular congestion and mild pulmonary edema, without evidence of pneumothorax or pleural effusion.

The significantly elevated BNP and pulmonary vascular congestion seen on chest x‐ray confirm the clinical diagnosis of heart failure. However, the negative troponin and unremarkable EKG suggest a non‐ischemic cause for her symptoms. An echocardiogram should be obtained with specific emphasis on the presence of valve regurgitation, pericardial effusion, and ventricular/atrial thickening consistent with infiltrative disorders. Thyroid stimulating hormone (TSH) and serologies for infectious agents, including, T. cruzi, HIV, CMV, and Toxoplasmosis gondii, should also be sent. The patient should receive intravenous loop diuretics to improve her cardiac dynamics and pulmonary edema.

Intravenous furosemide was administered. Her symptoms improved and oxygen saturation on room air was 92%. An echocardiogram revealed global hypokinesis with left ventricular ejection fraction (LVEF) of 35% to 40%. There was no evidence of an underlying valvular or infiltrative process. TSH was normal. T. cruzi antibodies were sent.

The echocardiogram did not reveal an underlying structural heart abnormality. Infiltrative cardiomyopathies do not typically demonstrate global hypokinesis on echocardiogram, particularly without evidence of ventricular wall thickening or increased echogenicity, that can be seen in amyloid and sarcoid cardiomyopathies. Therefore, infiltrative cardiomyopathy is unlikely to be a cause of this patient's heart failure. The rapid improvement of her symptoms with furosemide decreases the likelihood of infectious causes for her acute decompensation. In reviewing the patient's history, she had developed severe chronic diarrhea associated with poor oral intake and a 20‐pound weight loss prior to hospitalization. These symptoms, along with a history of osteoporosis at an early age without traditional risk factors, indicate a state of severe malnutrition, placing her at risk for thiamine deficiency. Checking the thiamine level would be appropriate.

Considering the patient's long history of malnutrition and negative infectious and ischemic evaluation, she was empirically treated for wet beriberi with thiamine supplementation through her TPN. A serum thiamine B₁ was obtained prior to supplementation. A vitamin D 25OH level was also sent, which was 15 ng/mL (normal >30 ng/mL), further suggesting malnutrition.

The patient continued to improve and furosemide was discontinued. Her initial serum thiamine level was 49 nmol/L (reference range: 70‐180 nmol/L). A repeat echocardiogram 5 days later revealed resolution of her systolic dysfunction and regional wall motion abnormality. The LVEF improved to 60%. Her colonoscopy biopsies revealed evidence of HIV enteropathy and CMV inclusion bodies. Her CMV viral load was 1223 genomes/mL. The T. cruzi antibodies were negative. She was restarted on HAART and ganciclovir. She continued to have diarrhea and was discharged home with TPN. Her serum thiamine level at discharge was 123 nmol/L.

Heart failure due to thiamine deficiency, or wet beriberi, was diagnosed considering the rapid clinical improvement in cardiac function after initiating thiamine therapy. While HIV cardiomyopathy could have contributed to heart failure in this patient, it is unlikely to improve so significantly over such a brief period of time.

DISCUSSION

Beriberi is a disease caused by severe thiamine deficiency. In fact, thiamine, also known as vitamin B₁, was first named the anti‐beriberi factor in 1926. However, the earliest descriptions of beriberi can be found in Chinese medical texts dating back to 2697 _BC.1 Beriberi is most commonly seen in Asia, where the diet is high in polished rice and the thiamine‐containing rice germs and husks have been removed. In the United States, thiamine‐enriched bread has virtually abolished the disease, except in severely malnourished populations such as alcoholics, those on fad diets, and patients with chronic diarrhea. Beriberi may also occur in patients with altered intestinal absorption such as post‐bariatric surgery patients.2 In 1985, the first case of beriberi as a complication of TPN without vitamin supplementation was reported.3 Subsequent cases of Wernicke's encephalopathy and beriberi have been noted in patients with gastrointestinal diseases and malabsorption on chronic TPN. More recently, thiamine deficiency has also been recognized in patients on long‐term diuretic therapy, as diuretics increase urinary excretion of this water‐soluble vitamin.4, 5 Since there is limited tissue storage of thiamine and its biologic half‐life is 10 to 20 days, high‐risk patients can develop thiamine deficiency within 4 weeks of initiation of diuretic therapy.6

Beriberi is classically divided into 2 types: wet, characterized by congestive heart failure, and dry, manifested as a symmetric peripheral neuropathy with both sensory and motor impairments.7 These 2 types of beriberi can coexist in the same patient; however, it is unclear why both types occur in some patients and not in others. Wet beriberi, also known as beriberi cardiomyopathy, typically presents as high‐output heart failure secondary to vasodilation, with a compensatory increase in blood volume and tachycardia.8 This state eventually leads to myocardial injury with systolic dysfunction and development of a low‐output state.8 Patients experience hypotension, lactic acidosis, and eventually fulminant vascular collapse. Although minor EKG changes such as sinus tachycardia, low‐voltage ventricular complexes, QT prolongation, and biphasic or inverted T waves are not uncommon in beriberi cardiomyopathy, major EKG changes, such as ST segment elevations and tall or deeply inverted T waves, are rare. Similarly, troponin elevation in beriberi cardiomyopathy is uncommon, but has been described.6

The pathogenesis of heart failure in beriberi is multifactorial. Thiamine is required for glucose to enter the Krebs cycle for aerobic metabolism, serving as a catalyst in the conversion of pyruvate to acetyl‐CoA. Without thiamine, anaerobic metabolism occurs, leading to the development of lactic acidosis and cellular malfunction. In fact, severe metabolic acidosis with serum pH values as low as 6.70 have been reported in cases of fulminant beriberi (although it is unclear if the lactic acidosis is mostly from anaerobic metabolism or from the low‐output state ultimately caused by thiamine deficiency).3

Laboratory diagnosis of thiamine deficiency, based on measurements of thiamine stores and metabolites, is often fraught with error and therefore unreliable. Serum pyruvate and lactate levels are commonly measured, and while elevated levels may be sensitive for thiamine deficiency, they are nonspecific. Measurement of whole blood thiamine is easy and the test is widely available; however, a low blood thiamine concentration is not always a sensitive indicator of deficiency since less than 1% of total body thiamine is found in whole blood.9 Additionally, this value may also be artificially elevated by thiamine intake immediately preceding the measurement. Urinary thiamine excretion has been proposed as a more accurate measurement, but this laboratory test is also problematic since urinary thiamine excretion reflects dietary intake more than total body stores.9 Erythrocyte transketolase activity (ETKA) is a functional enzyme test in which transketolase uses thiamine pyrophosphate as a catalyzer. This may be a more reliable measurement since red blood cells are among the first cells to be affected by thiamine depletion.9 Although a low ETKA level often indicates thiamine deficiency, this test is influenced by the hemoglobin concentration, and it is not widely available. Thus, the diagnosis of wet beriberi is usually made on the basis of rapid response to thiamine replacement.

Similar to the patient discussed, the clinical improvement in wet beriberi occurs within hours of treatment. There is an initial elevation in blood pressure and resolution of acidosis, followed by decrease in heart rate and normalization of cardiac output. Overall cardiac function improves within 24 to 48 hours after treatment, and return to a normal hemodynamic condition often occurs within 2 weeks of the start of treatment.10

There are no well‐established guidelines for the treatment of patients with beriberi, but general recommendations are an initial loading dose of intravenous thiamine 100 to 500 mg followed by 25 to 100 mg orally for 7 to 14 days.1 Thereafter, the daily thiamine requirement can be calculated based upon total caloric intake. The current recommendations in the United States are 0.5 mg of thiamine per 1000 kcal.1 However, one must consider whether a patient has impaired intestinal absorption or increased urinary losses when determining an appropriate maintenance dose.

Chronic malnutrition can lead to significant morbidity and mortality. Prior to admission, this patient already exhibited signs of severe malnutrition with a history of multiple pathologic fractures and diagnosis of osteoporosis. Considering her age and lack of risk factors for bone disease, osteoporosis suggests vitamin D deficiency. In this patient with chronic diarrhea caused by CMV, it is unlikely that a selective absorptive deficiency would occur. When the common causes of acute heart failure following volume challenge were excluded, the diagnosis of thiamine deficiency became more likely. Fortunately, an empiric trial of intravenous thiamine resulted in diagnosis by treatment and improvement of her cardiac function.

Ms Chen acutely worse, altered, please assist, room 522Beth, chirped my pager. Ever increasing time pressures meant that hospitalists were supervising rounds almost daily. I had sent my resident, Beth, and the rest of the team to round separately that day, to foster their independence. It looked like we would be meeting ahead of schedule.

I'd received a similar page 2 years earlier when I was a junior resident myself. From the beginning of internship, our faculty never hesitated to challenge us. I will never forget when one of the hospitalists who had just come across an unresponsive patient tapped me on the shoulder and casually asked, Hey, you wanna run a code? and will never forget my inadequacy or the specific assistance I required in those tense few minutes. He, and the ICU team that arrived, gave me every chance to lead, and supported me each time I hesitated.

In similar fashion, I had sent my intern, David, to admit a patient with suspected CHF. I received his urgent update shortly after our patient arrived on the cardiology floor: Mr Johnson dropping sats, please help, room 207. I jogged to the patient's room, where I found David, 3 nurses, 2 medical students, and in the center, Mr Johnson: lethargic, gray, cachectic, and making no effort to rise from the 40 degree incline of his hospital bed. Weak respirations fogged his non‐rebreather mask about 28 times a minute.

David offered a quick report: 74‐year‐old male, CAD, hypertension, dementia CHF exacerbation hypertensive to 190. I think he needs IV nitroglycerin and another 80 of lasix.

I was pleased to hear him commit to a diagnosis and plan, but after sitting Mr Johnson up for a quick exam, I couldn't agree. Are you sure? He sounds more junky than crackly. Neck veins are flat.

His EF is 25% and he's been here 3 times with CHF.

Well, that won't protect him from anything else. Mr Johnson slumped forward, accessory muscles firing weakly, and only half‐opened his eyes to a loud command and vigorous shake. Well, let's get the diagnosis later, what does he need, now?

Well, the lasix and the nitro

Assuming this is CHF, looking at him now, will that work fast enough to prevent intubation? David shook his head no. He's full code, right? Let's just call a code before he gets any worse. Anyone disagree? A nurse made the call, then guarded the door to turn away everyone but anesthesia and the MICU as they arrived.

So what do you think it is? David asked.

This doesn't smell like failure. He's not anxious, he's more obtunded than dyspneic. He looks hypercarbic. He doesn't have COPD?

Nah, just vomiting, then weaker, more confused, restless.

Maybe he aspirated. We'll see. So what do you want to have ready for anesthesia?

Um, meds. An IV. Chest X‐ray ready.

Good they bring the meds he's got an IV how about we pull the bed from the wall and raise it up, get some suction ready, take the headboard off? Nurses sprang into action.

If he's hypercarbic, shouldn't we bag him? David asked.

Good point, I said. David took the mask from the bag of emergency gear from the wall and started to fit it on Mr Johnson. It's a 2‐person job, if you want to hold the mask2 hands, good. A nurse began ventilations, and I added some cricoid pressure. Keeps us from inflating his stomach.

Seconds later, anesthesia arrived, and David provided a concise, organized summary. Mr Johnson was intubated and whisked without incident to the MICU, where bronchoscopy extracted several mucus plugs. He was soon extubated, and later recovered from a delirium which began with promethazine for nausea. It was the last year before the 80‐hour workweek regulations, and not once in the entire processfrom admission, to emergency on the ward, to initial MICU managementdid I or my fellow residents think to call an attending, although I'm sure we would have learned something, as I hadn't suspected a mucous plug. We weren't hiding anything. We were just taking care of our patient.

Two years later, it didn't seem odd that my junior resident called me for assistance with Ms Cheninitially. In room 522, much as I found Mr Johnson, I found Ms Chen: elderly, lethargic, gray, frail, laboring to breathe, rhythmically fogging a non‐rebreather mask 30 times a minute, only half‐opening her eyes to a vigorous shake. It was day 4 of her fifth hospitalization for bronchiectasis‐related respiratory failure within 2 months.

She just got a treatment but she still sounds awful, offered Beth. Indeed, Ms Chen's chest was gurgly and wheezy throughout. We put her on a non‐rebreather, but that hasn't helped.

I glanced at her monitor. Sat's 99%. What was she before?

96%.

So hypoxia isn't the problemwho's this? I asked, as transportation staff arrived.

Stat head CT for Chen, he replied.

I'm sorry, she can't go off the floor right now. Thanks for coming, I apologized, and sent him away. Beth, can you lay her flat or send her off the unit right now?

She's altered and I need to rule out stroke.

Let's talk about that later. I did a quick neuro exam as I spoke: Besides, she resists weak but equal; pupils and face symmetricshe's not focal. What's a more likely cause?

Metabolic? We can repeat her morning labs

Will they be different? Why is she here? What's her exam telling you?

Beth took in the scene before her, as Ms Chen struggled weakly to ventilate her lungs, and after a brief pause she had it worked out. She's hypercarbic. She needs an ABG. You think she plugged? She shook her head, and grasped Ms Chen's hand in her own. But she really hates suctioning.

Well, she's DNI, and without it, she could die. Beth agreed; we also called for noninvasive ventilation. But the team missed much of the action. The medical student missed the entire eventaside from attempting to summarize it from second‐hand reports for rounds the following day. I realized only later that her intern had been pushed to the back of the room for the critical decisions (much like the students during Mr Johnson's emergency), and headed out midway to attend a mandatory teaching sessionthe chief residents had begun taking attendance. The resident soon left for noon conference and afternoon clinic, enlisting me to write transfer orders and call the family. Finished with her other work, and under pressure to bank time against work hour limitations, which she was at risk of violating, the intern signed her pager over to me and left in the early afternoon, after sheepishly asking me if I wouldn't mind keeping an eye on our patient.

Later, a translator and I met with the Chens to comfort them and plan care for the family matriarch, having found a quiet solarium we could use, with summery views of the city and ocean in the distance to belie the grim topic of discussion.

What is your understanding of her lung problem right now?

Nay yeega jee um'jee huigor fai ho jing yeung?

What were her hopes and fears about her health?

Nay jee um'jee huigor see seung hai mai ho tai hoi?

My mind drifted during the Cantonese as I thought about how I use the unique teaching opportunities offered by wholly translated meetings. Never check the time. This body language says I am listening. I am speaking to them, not the translator. I make notes because families don't remember much after the C‐word, I would whisper to trainees while families conversed with translators. Now, as I began to discuss hospice philosophy, I felt acutely alone.

My team had missed most of a great hospital medicine experience: applying knowledge to manage a physiologic crisis; using communication skills to ease the resulting human crisis. Recently, to manage the latest set of work hour restrictions, our residency program withdrew from medicine consultation at 2 of 3 sites, and from the medicine wards at the hospital that serves most of our insured, geriatric, and oncology patients. The cost of this experiment to the overall residency experience is unknown. But cases like Ms Chen's remind me how much I missed being the primary doctor. I do not mind the new tasks I perform for my trainees. But I worry about what they are missing: sufficient responsibility for making key clinical decisions while protected by supervision on demand. I am glad my internship challenged meit prepared me for residency, moonlighting, and attending positions. Without a doubt, residency remains challenging, but it seems that the greatestor firstchallenge imposed on residents is now to beat the clock, not to become a well‐rounded, capable, independent physician.

That night, I complained to my spouse, then a psychiatry intern: We weren't giving our trainees the best preparation for a career in medicine the lengthy shift I spent managing a hypotensive crisis would be forbidden now my pre‐work hours interns were much happier than their work hours successors a 4‐year residency must be around the corner. The response I got was more bemused smile than grave concern. You don't think that's important? I asked.

Of course I do. It's just that with all this talk about the days of the giants, he said gently, you're starting to sound like a grumpy old man. We chuckled. He was right. I expect a lot from myself, my trainees, and every clinician. I'd figured I'd be worthy of the title at some point.

But at 30?

Ms Chen acutely worse, altered, please assist, room 522Beth, chirped my pager. Ever increasing time pressures meant that hospitalists were supervising rounds almost daily. I had sent my resident, Beth, and the rest of the team to round separately that day, to foster their independence. It looked like we would be meeting ahead of schedule.

I'd received a similar page 2 years earlier when I was a junior resident myself. From the beginning of internship, our faculty never hesitated to challenge us. I will never forget when one of the hospitalists who had just come across an unresponsive patient tapped me on the shoulder and casually asked, Hey, you wanna run a code? and will never forget my inadequacy or the specific assistance I required in those tense few minutes. He, and the ICU team that arrived, gave me every chance to lead, and supported me each time I hesitated.

In similar fashion, I had sent my intern, David, to admit a patient with suspected CHF. I received his urgent update shortly after our patient arrived on the cardiology floor: Mr Johnson dropping sats, please help, room 207. I jogged to the patient's room, where I found David, 3 nurses, 2 medical students, and in the center, Mr Johnson: lethargic, gray, cachectic, and making no effort to rise from the 40 degree incline of his hospital bed. Weak respirations fogged his non‐rebreather mask about 28 times a minute.

David offered a quick report: 74‐year‐old male, CAD, hypertension, dementia CHF exacerbation hypertensive to 190. I think he needs IV nitroglycerin and another 80 of lasix.

I was pleased to hear him commit to a diagnosis and plan, but after sitting Mr Johnson up for a quick exam, I couldn't agree. Are you sure? He sounds more junky than crackly. Neck veins are flat.

His EF is 25% and he's been here 3 times with CHF.

Well, that won't protect him from anything else. Mr Johnson slumped forward, accessory muscles firing weakly, and only half‐opened his eyes to a loud command and vigorous shake. Well, let's get the diagnosis later, what does he need, now?

Well, the lasix and the nitro

Assuming this is CHF, looking at him now, will that work fast enough to prevent intubation? David shook his head no. He's full code, right? Let's just call a code before he gets any worse. Anyone disagree? A nurse made the call, then guarded the door to turn away everyone but anesthesia and the MICU as they arrived.

So what do you think it is? David asked.

This doesn't smell like failure. He's not anxious, he's more obtunded than dyspneic. He looks hypercarbic. He doesn't have COPD?

Nah, just vomiting, then weaker, more confused, restless.

Maybe he aspirated. We'll see. So what do you want to have ready for anesthesia?

Um, meds. An IV. Chest X‐ray ready.

Good they bring the meds he's got an IV how about we pull the bed from the wall and raise it up, get some suction ready, take the headboard off? Nurses sprang into action.

If he's hypercarbic, shouldn't we bag him? David asked.

Good point, I said. David took the mask from the bag of emergency gear from the wall and started to fit it on Mr Johnson. It's a 2‐person job, if you want to hold the mask2 hands, good. A nurse began ventilations, and I added some cricoid pressure. Keeps us from inflating his stomach.

Seconds later, anesthesia arrived, and David provided a concise, organized summary. Mr Johnson was intubated and whisked without incident to the MICU, where bronchoscopy extracted several mucus plugs. He was soon extubated, and later recovered from a delirium which began with promethazine for nausea. It was the last year before the 80‐hour workweek regulations, and not once in the entire processfrom admission, to emergency on the ward, to initial MICU managementdid I or my fellow residents think to call an attending, although I'm sure we would have learned something, as I hadn't suspected a mucous plug. We weren't hiding anything. We were just taking care of our patient.

Two years later, it didn't seem odd that my junior resident called me for assistance with Ms Cheninitially. In room 522, much as I found Mr Johnson, I found Ms Chen: elderly, lethargic, gray, frail, laboring to breathe, rhythmically fogging a non‐rebreather mask 30 times a minute, only half‐opening her eyes to a vigorous shake. It was day 4 of her fifth hospitalization for bronchiectasis‐related respiratory failure within 2 months.

She just got a treatment but she still sounds awful, offered Beth. Indeed, Ms Chen's chest was gurgly and wheezy throughout. We put her on a non‐rebreather, but that hasn't helped.

I glanced at her monitor. Sat's 99%. What was she before?

96%.

So hypoxia isn't the problemwho's this? I asked, as transportation staff arrived.

Stat head CT for Chen, he replied.

I'm sorry, she can't go off the floor right now. Thanks for coming, I apologized, and sent him away. Beth, can you lay her flat or send her off the unit right now?

She's altered and I need to rule out stroke.

Let's talk about that later. I did a quick neuro exam as I spoke: Besides, she resists weak but equal; pupils and face symmetricshe's not focal. What's a more likely cause?

Metabolic? We can repeat her morning labs

Will they be different? Why is she here? What's her exam telling you?

Beth took in the scene before her, as Ms Chen struggled weakly to ventilate her lungs, and after a brief pause she had it worked out. She's hypercarbic. She needs an ABG. You think she plugged? She shook her head, and grasped Ms Chen's hand in her own. But she really hates suctioning.

Well, she's DNI, and without it, she could die. Beth agreed; we also called for noninvasive ventilation. But the team missed much of the action. The medical student missed the entire eventaside from attempting to summarize it from second‐hand reports for rounds the following day. I realized only later that her intern had been pushed to the back of the room for the critical decisions (much like the students during Mr Johnson's emergency), and headed out midway to attend a mandatory teaching sessionthe chief residents had begun taking attendance. The resident soon left for noon conference and afternoon clinic, enlisting me to write transfer orders and call the family. Finished with her other work, and under pressure to bank time against work hour limitations, which she was at risk of violating, the intern signed her pager over to me and left in the early afternoon, after sheepishly asking me if I wouldn't mind keeping an eye on our patient.

Later, a translator and I met with the Chens to comfort them and plan care for the family matriarch, having found a quiet solarium we could use, with summery views of the city and ocean in the distance to belie the grim topic of discussion.

What is your understanding of her lung problem right now?

Nay yeega jee um'jee huigor fai ho jing yeung?

What were her hopes and fears about her health?

Nay jee um'jee huigor see seung hai mai ho tai hoi?

My mind drifted during the Cantonese as I thought about how I use the unique teaching opportunities offered by wholly translated meetings. Never check the time. This body language says I am listening. I am speaking to them, not the translator. I make notes because families don't remember much after the C‐word, I would whisper to trainees while families conversed with translators. Now, as I began to discuss hospice philosophy, I felt acutely alone.

My team had missed most of a great hospital medicine experience: applying knowledge to manage a physiologic crisis; using communication skills to ease the resulting human crisis. Recently, to manage the latest set of work hour restrictions, our residency program withdrew from medicine consultation at 2 of 3 sites, and from the medicine wards at the hospital that serves most of our insured, geriatric, and oncology patients. The cost of this experiment to the overall residency experience is unknown. But cases like Ms Chen's remind me how much I missed being the primary doctor. I do not mind the new tasks I perform for my trainees. But I worry about what they are missing: sufficient responsibility for making key clinical decisions while protected by supervision on demand. I am glad my internship challenged meit prepared me for residency, moonlighting, and attending positions. Without a doubt, residency remains challenging, but it seems that the greatestor firstchallenge imposed on residents is now to beat the clock, not to become a well‐rounded, capable, independent physician.

That night, I complained to my spouse, then a psychiatry intern: We weren't giving our trainees the best preparation for a career in medicine the lengthy shift I spent managing a hypotensive crisis would be forbidden now my pre‐work hours interns were much happier than their work hours successors a 4‐year residency must be around the corner. The response I got was more bemused smile than grave concern. You don't think that's important? I asked.

Of course I do. It's just that with all this talk about the days of the giants, he said gently, you're starting to sound like a grumpy old man. We chuckled. He was right. I expect a lot from myself, my trainees, and every clinician. I'd figured I'd be worthy of the title at some point.

But at 30?

Ms Chen acutely worse, altered, please assist, room 522Beth, chirped my pager. Ever increasing time pressures meant that hospitalists were supervising rounds almost daily. I had sent my resident, Beth, and the rest of the team to round separately that day, to foster their independence. It looked like we would be meeting ahead of schedule.

I'd received a similar page 2 years earlier when I was a junior resident myself. From the beginning of internship, our faculty never hesitated to challenge us. I will never forget when one of the hospitalists who had just come across an unresponsive patient tapped me on the shoulder and casually asked, Hey, you wanna run a code? and will never forget my inadequacy or the specific assistance I required in those tense few minutes. He, and the ICU team that arrived, gave me every chance to lead, and supported me each time I hesitated.

In similar fashion, I had sent my intern, David, to admit a patient with suspected CHF. I received his urgent update shortly after our patient arrived on the cardiology floor: Mr Johnson dropping sats, please help, room 207. I jogged to the patient's room, where I found David, 3 nurses, 2 medical students, and in the center, Mr Johnson: lethargic, gray, cachectic, and making no effort to rise from the 40 degree incline of his hospital bed. Weak respirations fogged his non‐rebreather mask about 28 times a minute.

David offered a quick report: 74‐year‐old male, CAD, hypertension, dementia CHF exacerbation hypertensive to 190. I think he needs IV nitroglycerin and another 80 of lasix.

I was pleased to hear him commit to a diagnosis and plan, but after sitting Mr Johnson up for a quick exam, I couldn't agree. Are you sure? He sounds more junky than crackly. Neck veins are flat.

His EF is 25% and he's been here 3 times with CHF.

Well, that won't protect him from anything else. Mr Johnson slumped forward, accessory muscles firing weakly, and only half‐opened his eyes to a loud command and vigorous shake. Well, let's get the diagnosis later, what does he need, now?

Well, the lasix and the nitro

Assuming this is CHF, looking at him now, will that work fast enough to prevent intubation? David shook his head no. He's full code, right? Let's just call a code before he gets any worse. Anyone disagree? A nurse made the call, then guarded the door to turn away everyone but anesthesia and the MICU as they arrived.

So what do you think it is? David asked.

This doesn't smell like failure. He's not anxious, he's more obtunded than dyspneic. He looks hypercarbic. He doesn't have COPD?

Nah, just vomiting, then weaker, more confused, restless.

Maybe he aspirated. We'll see. So what do you want to have ready for anesthesia?

Um, meds. An IV. Chest X‐ray ready.

Good they bring the meds he's got an IV how about we pull the bed from the wall and raise it up, get some suction ready, take the headboard off? Nurses sprang into action.

If he's hypercarbic, shouldn't we bag him? David asked.

Good point, I said. David took the mask from the bag of emergency gear from the wall and started to fit it on Mr Johnson. It's a 2‐person job, if you want to hold the mask2 hands, good. A nurse began ventilations, and I added some cricoid pressure. Keeps us from inflating his stomach.

Seconds later, anesthesia arrived, and David provided a concise, organized summary. Mr Johnson was intubated and whisked without incident to the MICU, where bronchoscopy extracted several mucus plugs. He was soon extubated, and later recovered from a delirium which began with promethazine for nausea. It was the last year before the 80‐hour workweek regulations, and not once in the entire processfrom admission, to emergency on the ward, to initial MICU managementdid I or my fellow residents think to call an attending, although I'm sure we would have learned something, as I hadn't suspected a mucous plug. We weren't hiding anything. We were just taking care of our patient.

Two years later, it didn't seem odd that my junior resident called me for assistance with Ms Cheninitially. In room 522, much as I found Mr Johnson, I found Ms Chen: elderly, lethargic, gray, frail, laboring to breathe, rhythmically fogging a non‐rebreather mask 30 times a minute, only half‐opening her eyes to a vigorous shake. It was day 4 of her fifth hospitalization for bronchiectasis‐related respiratory failure within 2 months.

She just got a treatment but she still sounds awful, offered Beth. Indeed, Ms Chen's chest was gurgly and wheezy throughout. We put her on a non‐rebreather, but that hasn't helped.

I glanced at her monitor. Sat's 99%. What was she before?

96%.

So hypoxia isn't the problemwho's this? I asked, as transportation staff arrived.

Stat head CT for Chen, he replied.

I'm sorry, she can't go off the floor right now. Thanks for coming, I apologized, and sent him away. Beth, can you lay her flat or send her off the unit right now?

She's altered and I need to rule out stroke.

Let's talk about that later. I did a quick neuro exam as I spoke: Besides, she resists weak but equal; pupils and face symmetricshe's not focal. What's a more likely cause?

Metabolic? We can repeat her morning labs

Will they be different? Why is she here? What's her exam telling you?

Beth took in the scene before her, as Ms Chen struggled weakly to ventilate her lungs, and after a brief pause she had it worked out. She's hypercarbic. She needs an ABG. You think she plugged? She shook her head, and grasped Ms Chen's hand in her own. But she really hates suctioning.

Well, she's DNI, and without it, she could die. Beth agreed; we also called for noninvasive ventilation. But the team missed much of the action. The medical student missed the entire eventaside from attempting to summarize it from second‐hand reports for rounds the following day. I realized only later that her intern had been pushed to the back of the room for the critical decisions (much like the students during Mr Johnson's emergency), and headed out midway to attend a mandatory teaching sessionthe chief residents had begun taking attendance. The resident soon left for noon conference and afternoon clinic, enlisting me to write transfer orders and call the family. Finished with her other work, and under pressure to bank time against work hour limitations, which she was at risk of violating, the intern signed her pager over to me and left in the early afternoon, after sheepishly asking me if I wouldn't mind keeping an eye on our patient.

Later, a translator and I met with the Chens to comfort them and plan care for the family matriarch, having found a quiet solarium we could use, with summery views of the city and ocean in the distance to belie the grim topic of discussion.

What is your understanding of her lung problem right now?

Nay yeega jee um'jee huigor fai ho jing yeung?

What were her hopes and fears about her health?

Nay jee um'jee huigor see seung hai mai ho tai hoi?

My mind drifted during the Cantonese as I thought about how I use the unique teaching opportunities offered by wholly translated meetings. Never check the time. This body language says I am listening. I am speaking to them, not the translator. I make notes because families don't remember much after the C‐word, I would whisper to trainees while families conversed with translators. Now, as I began to discuss hospice philosophy, I felt acutely alone.

My team had missed most of a great hospital medicine experience: applying knowledge to manage a physiologic crisis; using communication skills to ease the resulting human crisis. Recently, to manage the latest set of work hour restrictions, our residency program withdrew from medicine consultation at 2 of 3 sites, and from the medicine wards at the hospital that serves most of our insured, geriatric, and oncology patients. The cost of this experiment to the overall residency experience is unknown. But cases like Ms Chen's remind me how much I missed being the primary doctor. I do not mind the new tasks I perform for my trainees. But I worry about what they are missing: sufficient responsibility for making key clinical decisions while protected by supervision on demand. I am glad my internship challenged meit prepared me for residency, moonlighting, and attending positions. Without a doubt, residency remains challenging, but it seems that the greatestor firstchallenge imposed on residents is now to beat the clock, not to become a well‐rounded, capable, independent physician.

That night, I complained to my spouse, then a psychiatry intern: We weren't giving our trainees the best preparation for a career in medicine the lengthy shift I spent managing a hypotensive crisis would be forbidden now my pre‐work hours interns were much happier than their work hours successors a 4‐year residency must be around the corner. The response I got was more bemused smile than grave concern. You don't think that's important? I asked.

Of course I do. It's just that with all this talk about the days of the giants, he said gently, you're starting to sound like a grumpy old man. We chuckled. He was right. I expect a lot from myself, my trainees, and every clinician. I'd figured I'd be worthy of the title at some point.

But at 30?

Enterococci have been identified as a causative organism in approximately 10% of all nosocomial bloodstream infections (BSIs).1, 2 In 2006, the Infectious Diseases Society of America identified vancomycin‐ resistant Enterococcus faecium (VRE) as 1 of 6 microbes considered to be among the most dangerous due to high rates of resistance and a limited number of effective antimicrobials.3 E. faecium has exhibited high rates of glycopeptide resistance with as many as 60% of isolates from BSIs being resistant to vancomycin.2, 4 Due to increasing resistance to glycopeptides, vancomycin has become obsolete in the treatment of E. faecium infections.5

A limited number of antimicrobials are available for the treatment of infections due to VRE. Agents active in vitro are quinupristin‐dalfopristin, tigecycline, linezolid, and daptomycin. Quinupristin‐dalfopristin was one of the first agents approved for use in VRE infections; however, treatment with this agent has been limited because of mediocre clinical response rates, undesirable adverse effects, high cost, and insufficient E. faecalis activity.6, 7 Tigecycline is not an optimal antibiotic for the treatment of VRE bacteremia, because serum concentrations achieved after administration are inadequate to treat BSIs.7 In contrast, linezolid and daptomycin have evinced efficacy against VRE bacteremia, with reported microbiologic response rates of 85% and 80%, respectively.7, 8 One inherent difference between these antibiotics that may theoretically affect their use in immunocompromised patients is that linezolid is bacteriostatic, whereas daptomycin is bactericidal. It has been postulated that by using a bactericidal antibiotic such as daptomycin in the immunocompromised host, one may achieve superior clinical and microbiologic response rates.3, 7, 9, 10

Since the introduction of the oxazolidinone linezolid in 2000, widespread use has led to reports of linezolid‐resistant VRE as well as nosocomial transmission of linezolid‐resistant VRE in hospitals.4, 1114 Despite linezolid being a key antibiotic for the treatment of VRE infections over the last 10 years, development of resistance along with potential hematologic and neurologic toxicity during long‐term use remains a concern.7, 8 Although daptomycin is active against several resistant organisms, including VRE, the evidence supporting use of daptomycin for VRE BSI is limited to case reports or small case series.7, 9 Moreover, daptomycin has not received US Food and Drug Administration approval for the treatment of VRE infections,15 and emerging data regarding daptomycin‐nonsusceptible enterococci (Minimum Inhibitory Concentration, MIC >4 mg/L) highlight a new problem for this multidrug‐resistant pathogen.16, 17 Few studies in recent years have compared these 2 antibiotics in the treatment of VRE BSIs.4, 18, 19 Due to the high rates of vancomycin resistance reported at our institution and the ubiquitous use of linezolid and daptomycin in the treatment of VRE bacteremia, we chose to evaluate response rates for these antibiotics in an effort to add to previously published literature on this topic.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Vergis et al21 reported that infections with VRE compared with vancomycin‐sensitive infections were associated with a higher rate of mortality and that the chosen antimicrobial therapy may play a pivotal role in the risk of death. Our retrospective study suggests that linezolid and daptomycin appear to be equally efficacious for the treatment of VRE BSIs. The results from our study for clinical and microbiologic cure rates for linezolid and daptomycin are similar to previously published data.7, 8 In accordance with previous studies,4 our data demonstrate that there is a higher rate of recurrence in patients treated with daptomycin. This finding may be explained by the fact that the daptomycin group was comprised of more complex patients with a greater disease burden versus the linezolid group; therefore, they were more susceptible to a recurrent VRE infection. In our study, patients who were treated with daptomycin were 5.5 times more likely to have a recurrent infection than linezolid‐treated patients. However, this finding must be scrutinized, because over half of the patients with recurrence either received an inappropriate dose or had high MICs to daptomycin.

Despite there being few clinical and microbiologic outcome data with daptomycin, our study proposes that a bactericidal antibiotic and a bacteriostatic antibiotic have comparable efficacy in the treatment of VRE BSIs. Previous literature has mainly comprised case studies or series that have evaluated clinical outcomes with daptomycin in the treatment of VRE BSIs. Gallagher et al7 reported the results of a retrospective case series of 30 patients with VRE bacteremia who were treated with daptomycin. In this study, microbiologic cure was achieved in 80% of patients, with clinical success in 59% of the patients. In 2009, Mave et al4 compared clinical outcomes between daptomycin and linezolid in the treatment of VRE bacteremia. Reported results demonstrated a microbiologic cure rate of 90% for daptomycin versus 88% for linezolid.4 Moreover, there were no differences in mortality between the groups in our study. In 2010, Crank et al18 reported no differences in mortality (in‐hospital) for hospitalized patients with VRE BSIs treated with linezolid or daptomycin. Our results seem to be consistent with what has been published previously concerning clinical outcomes associated with linezolid or daptomycin in the treatment of VRE BSI.

The average daptomycin dose received in our patients was 6.1 mg/kg with doses ranging from 3.410.4 mg/kg. The underdosing as well as higher MICs to daptomycin may have contributed to a higher rate of recurrence. Previous reports state that Enterococcus species may have higher MICs to daptomycin than Staphylococcus or Streptococcus species; consequently, higher doses may be needed to adequately treat enterococcal infections.7 In the aforementioned study by Gallagher et al,7 doses of daptomycin 6 mg/kg were associated with a positive clinical outcome in 81% of patients compared with 31% if the dose used was <6 mg/kg. Linezolid is dosed 600 mg every 12 hours by mouth or intravenously, with no variations. There have been no studies comparing the uniform dosing of linezolid to the weight‐based dosing of daptomycin and their effects on outcomes.

Patients particularly susceptible to VRE infections include those with neutropenia and/or cancer, patients receiving long‐term hemodialysis, and liver transplant recipients.3, 22, 23 Upon review of this immunocompromised population, we noted no statistically significant differences in overall outcomes. A study by Kraft et al.24 supports the findings in our study that both drugs appear useful in the treatment of VRE bacteremia in patients with hematologic malignancy. We did identify a difference, albeit nonsignificant, in LOS for daptomycin versus linezolid in patients with a history of liver transplantation. Again, the level of care that these patients needed compared with the general population may explain this difference. As mentioned previously, another pertinent factor would be the dose of daptomycin used in these patients, because the dose can affect clinical success. Because all of the other patients had a similar LOS, we cannot determine that the increased LOS seen in liver transplant patients treated with daptomycin was solely due to daptomycin use. The reason for the increased LOS seems to be multifactorial. In the neutropenic population, a difference in LOS was also recognized, but follow‐up complete blood count values were not collected for these patients to determine whether linezolid contributed to further bone marrow suppression leading to an increase in LOS. For both of these patient populations, the number of patients included is very small (n = 21 for neutropenia total, n = 13 for liver transplant total), which can lead to a high degree of variance.

This study has several limitations. This was a retrospective review; therefore, we had no control over the selection of therapy. This may be reflected in an apparent preferential use of daptomycin in immunocompromised patients. Furthermore, 62% of linezolid patients and 54% of daptomycin patients received an antibiotic before initial therapy that could have potentially altered response rates. Due to the paucity of documentation surrounding initial site of infection, some of the positive cultures may represent potential contamination, because VRE may contaminate skin.25 Contamination seems implausible, however, because patients were seen by an infectious disease physician and had at least 1 documented positive VRE blood culture. We chose arbitrary definitions for clinical cure, microbiologic cure, microbiologic failure, recurrence, and reinfection. Previous studies have used their own definitions leading to discrepancies in reporting. Another limitation was that follow‐up cultures were not obtained on all of the patients, which was needed to determine rates of recurrence, reinfection, and microbiologic cure. MICs to daptomycin were not reported in 30% of our patients, potentially altering the recurrence rate seen in the daptomycin‐treated patients. Because clinical cure was not documented in the chart, it was inferred from the laboratory values and vital sign information. One investigator analyzed all of the values and made the determination of clinical cure, allowing for a consistent approach to data review.

In the face of the imposing threat of a highly resistant organism such as VRE with a limited number of efficacious antibiotics, antimicrobial selection becomes increasingly important and is requisite to clinical and microbiological success. To our knowledge, this is one of the largest studies to date comparing the efficacy of linezolid with that of daptomycin in the treatment of VRE bacteremia. Both of these agents are effective for the treatment of VRE BSIs. Nevertheless, specific factors related to the medication (eg, dose, route of administration) as well as the patient (eg, comorbid conditions, acuity of illness) should be taken into consideration when selecting an initial antimicrobial agent. Because the treatment of VRE BSIs continues to be a challenge, larger prospective randomized controlled trials are needed to corroborate our results and determine the optimal therapy for this serious infection.

Enterococci have been identified as a causative organism in approximately 10% of all nosocomial bloodstream infections (BSIs).1, 2 In 2006, the Infectious Diseases Society of America identified vancomycin‐ resistant Enterococcus faecium (VRE) as 1 of 6 microbes considered to be among the most dangerous due to high rates of resistance and a limited number of effective antimicrobials.3 E. faecium has exhibited high rates of glycopeptide resistance with as many as 60% of isolates from BSIs being resistant to vancomycin.2, 4 Due to increasing resistance to glycopeptides, vancomycin has become obsolete in the treatment of E. faecium infections.5

A limited number of antimicrobials are available for the treatment of infections due to VRE. Agents active in vitro are quinupristin‐dalfopristin, tigecycline, linezolid, and daptomycin. Quinupristin‐dalfopristin was one of the first agents approved for use in VRE infections; however, treatment with this agent has been limited because of mediocre clinical response rates, undesirable adverse effects, high cost, and insufficient E. faecalis activity.6, 7 Tigecycline is not an optimal antibiotic for the treatment of VRE bacteremia, because serum concentrations achieved after administration are inadequate to treat BSIs.7 In contrast, linezolid and daptomycin have evinced efficacy against VRE bacteremia, with reported microbiologic response rates of 85% and 80%, respectively.7, 8 One inherent difference between these antibiotics that may theoretically affect their use in immunocompromised patients is that linezolid is bacteriostatic, whereas daptomycin is bactericidal. It has been postulated that by using a bactericidal antibiotic such as daptomycin in the immunocompromised host, one may achieve superior clinical and microbiologic response rates.3, 7, 9, 10

Since the introduction of the oxazolidinone linezolid in 2000, widespread use has led to reports of linezolid‐resistant VRE as well as nosocomial transmission of linezolid‐resistant VRE in hospitals.4, 1114 Despite linezolid being a key antibiotic for the treatment of VRE infections over the last 10 years, development of resistance along with potential hematologic and neurologic toxicity during long‐term use remains a concern.7, 8 Although daptomycin is active against several resistant organisms, including VRE, the evidence supporting use of daptomycin for VRE BSI is limited to case reports or small case series.7, 9 Moreover, daptomycin has not received US Food and Drug Administration approval for the treatment of VRE infections,15 and emerging data regarding daptomycin‐nonsusceptible enterococci (Minimum Inhibitory Concentration, MIC >4 mg/L) highlight a new problem for this multidrug‐resistant pathogen.16, 17 Few studies in recent years have compared these 2 antibiotics in the treatment of VRE BSIs.4, 18, 19 Due to the high rates of vancomycin resistance reported at our institution and the ubiquitous use of linezolid and daptomycin in the treatment of VRE bacteremia, we chose to evaluate response rates for these antibiotics in an effort to add to previously published literature on this topic.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Vergis et al21 reported that infections with VRE compared with vancomycin‐sensitive infections were associated with a higher rate of mortality and that the chosen antimicrobial therapy may play a pivotal role in the risk of death. Our retrospective study suggests that linezolid and daptomycin appear to be equally efficacious for the treatment of VRE BSIs. The results from our study for clinical and microbiologic cure rates for linezolid and daptomycin are similar to previously published data.7, 8 In accordance with previous studies,4 our data demonstrate that there is a higher rate of recurrence in patients treated with daptomycin. This finding may be explained by the fact that the daptomycin group was comprised of more complex patients with a greater disease burden versus the linezolid group; therefore, they were more susceptible to a recurrent VRE infection. In our study, patients who were treated with daptomycin were 5.5 times more likely to have a recurrent infection than linezolid‐treated patients. However, this finding must be scrutinized, because over half of the patients with recurrence either received an inappropriate dose or had high MICs to daptomycin.

Despite there being few clinical and microbiologic outcome data with daptomycin, our study proposes that a bactericidal antibiotic and a bacteriostatic antibiotic have comparable efficacy in the treatment of VRE BSIs. Previous literature has mainly comprised case studies or series that have evaluated clinical outcomes with daptomycin in the treatment of VRE BSIs. Gallagher et al7 reported the results of a retrospective case series of 30 patients with VRE bacteremia who were treated with daptomycin. In this study, microbiologic cure was achieved in 80% of patients, with clinical success in 59% of the patients. In 2009, Mave et al4 compared clinical outcomes between daptomycin and linezolid in the treatment of VRE bacteremia. Reported results demonstrated a microbiologic cure rate of 90% for daptomycin versus 88% for linezolid.4 Moreover, there were no differences in mortality between the groups in our study. In 2010, Crank et al18 reported no differences in mortality (in‐hospital) for hospitalized patients with VRE BSIs treated with linezolid or daptomycin. Our results seem to be consistent with what has been published previously concerning clinical outcomes associated with linezolid or daptomycin in the treatment of VRE BSI.

The average daptomycin dose received in our patients was 6.1 mg/kg with doses ranging from 3.410.4 mg/kg. The underdosing as well as higher MICs to daptomycin may have contributed to a higher rate of recurrence. Previous reports state that Enterococcus species may have higher MICs to daptomycin than Staphylococcus or Streptococcus species; consequently, higher doses may be needed to adequately treat enterococcal infections.7 In the aforementioned study by Gallagher et al,7 doses of daptomycin 6 mg/kg were associated with a positive clinical outcome in 81% of patients compared with 31% if the dose used was <6 mg/kg. Linezolid is dosed 600 mg every 12 hours by mouth or intravenously, with no variations. There have been no studies comparing the uniform dosing of linezolid to the weight‐based dosing of daptomycin and their effects on outcomes.

Patients particularly susceptible to VRE infections include those with neutropenia and/or cancer, patients receiving long‐term hemodialysis, and liver transplant recipients.3, 22, 23 Upon review of this immunocompromised population, we noted no statistically significant differences in overall outcomes. A study by Kraft et al.24 supports the findings in our study that both drugs appear useful in the treatment of VRE bacteremia in patients with hematologic malignancy. We did identify a difference, albeit nonsignificant, in LOS for daptomycin versus linezolid in patients with a history of liver transplantation. Again, the level of care that these patients needed compared with the general population may explain this difference. As mentioned previously, another pertinent factor would be the dose of daptomycin used in these patients, because the dose can affect clinical success. Because all of the other patients had a similar LOS, we cannot determine that the increased LOS seen in liver transplant patients treated with daptomycin was solely due to daptomycin use. The reason for the increased LOS seems to be multifactorial. In the neutropenic population, a difference in LOS was also recognized, but follow‐up complete blood count values were not collected for these patients to determine whether linezolid contributed to further bone marrow suppression leading to an increase in LOS. For both of these patient populations, the number of patients included is very small (n = 21 for neutropenia total, n = 13 for liver transplant total), which can lead to a high degree of variance.

This study has several limitations. This was a retrospective review; therefore, we had no control over the selection of therapy. This may be reflected in an apparent preferential use of daptomycin in immunocompromised patients. Furthermore, 62% of linezolid patients and 54% of daptomycin patients received an antibiotic before initial therapy that could have potentially altered response rates. Due to the paucity of documentation surrounding initial site of infection, some of the positive cultures may represent potential contamination, because VRE may contaminate skin.25 Contamination seems implausible, however, because patients were seen by an infectious disease physician and had at least 1 documented positive VRE blood culture. We chose arbitrary definitions for clinical cure, microbiologic cure, microbiologic failure, recurrence, and reinfection. Previous studies have used their own definitions leading to discrepancies in reporting. Another limitation was that follow‐up cultures were not obtained on all of the patients, which was needed to determine rates of recurrence, reinfection, and microbiologic cure. MICs to daptomycin were not reported in 30% of our patients, potentially altering the recurrence rate seen in the daptomycin‐treated patients. Because clinical cure was not documented in the chart, it was inferred from the laboratory values and vital sign information. One investigator analyzed all of the values and made the determination of clinical cure, allowing for a consistent approach to data review.

In the face of the imposing threat of a highly resistant organism such as VRE with a limited number of efficacious antibiotics, antimicrobial selection becomes increasingly important and is requisite to clinical and microbiological success. To our knowledge, this is one of the largest studies to date comparing the efficacy of linezolid with that of daptomycin in the treatment of VRE bacteremia. Both of these agents are effective for the treatment of VRE BSIs. Nevertheless, specific factors related to the medication (eg, dose, route of administration) as well as the patient (eg, comorbid conditions, acuity of illness) should be taken into consideration when selecting an initial antimicrobial agent. Because the treatment of VRE BSIs continues to be a challenge, larger prospective randomized controlled trials are needed to corroborate our results and determine the optimal therapy for this serious infection.

Enterococci have been identified as a causative organism in approximately 10% of all nosocomial bloodstream infections (BSIs).1, 2 In 2006, the Infectious Diseases Society of America identified vancomycin‐ resistant Enterococcus faecium (VRE) as 1 of 6 microbes considered to be among the most dangerous due to high rates of resistance and a limited number of effective antimicrobials.3 E. faecium has exhibited high rates of glycopeptide resistance with as many as 60% of isolates from BSIs being resistant to vancomycin.2, 4 Due to increasing resistance to glycopeptides, vancomycin has become obsolete in the treatment of E. faecium infections.5

A limited number of antimicrobials are available for the treatment of infections due to VRE. Agents active in vitro are quinupristin‐dalfopristin, tigecycline, linezolid, and daptomycin. Quinupristin‐dalfopristin was one of the first agents approved for use in VRE infections; however, treatment with this agent has been limited because of mediocre clinical response rates, undesirable adverse effects, high cost, and insufficient E. faecalis activity.6, 7 Tigecycline is not an optimal antibiotic for the treatment of VRE bacteremia, because serum concentrations achieved after administration are inadequate to treat BSIs.7 In contrast, linezolid and daptomycin have evinced efficacy against VRE bacteremia, with reported microbiologic response rates of 85% and 80%, respectively.7, 8 One inherent difference between these antibiotics that may theoretically affect their use in immunocompromised patients is that linezolid is bacteriostatic, whereas daptomycin is bactericidal. It has been postulated that by using a bactericidal antibiotic such as daptomycin in the immunocompromised host, one may achieve superior clinical and microbiologic response rates.3, 7, 9, 10

Since the introduction of the oxazolidinone linezolid in 2000, widespread use has led to reports of linezolid‐resistant VRE as well as nosocomial transmission of linezolid‐resistant VRE in hospitals.4, 1114 Despite linezolid being a key antibiotic for the treatment of VRE infections over the last 10 years, development of resistance along with potential hematologic and neurologic toxicity during long‐term use remains a concern.7, 8 Although daptomycin is active against several resistant organisms, including VRE, the evidence supporting use of daptomycin for VRE BSI is limited to case reports or small case series.7, 9 Moreover, daptomycin has not received US Food and Drug Administration approval for the treatment of VRE infections,15 and emerging data regarding daptomycin‐nonsusceptible enterococci (Minimum Inhibitory Concentration, MIC >4 mg/L) highlight a new problem for this multidrug‐resistant pathogen.16, 17 Few studies in recent years have compared these 2 antibiotics in the treatment of VRE BSIs.4, 18, 19 Due to the high rates of vancomycin resistance reported at our institution and the ubiquitous use of linezolid and daptomycin in the treatment of VRE bacteremia, we chose to evaluate response rates for these antibiotics in an effort to add to previously published literature on this topic.

MATERIALS AND METHODS

RESULTS

DISCUSSION

Vergis et al21 reported that infections with VRE compared with vancomycin‐sensitive infections were associated with a higher rate of mortality and that the chosen antimicrobial therapy may play a pivotal role in the risk of death. Our retrospective study suggests that linezolid and daptomycin appear to be equally efficacious for the treatment of VRE BSIs. The results from our study for clinical and microbiologic cure rates for linezolid and daptomycin are similar to previously published data.7, 8 In accordance with previous studies,4 our data demonstrate that there is a higher rate of recurrence in patients treated with daptomycin. This finding may be explained by the fact that the daptomycin group was comprised of more complex patients with a greater disease burden versus the linezolid group; therefore, they were more susceptible to a recurrent VRE infection. In our study, patients who were treated with daptomycin were 5.5 times more likely to have a recurrent infection than linezolid‐treated patients. However, this finding must be scrutinized, because over half of the patients with recurrence either received an inappropriate dose or had high MICs to daptomycin.

Despite there being few clinical and microbiologic outcome data with daptomycin, our study proposes that a bactericidal antibiotic and a bacteriostatic antibiotic have comparable efficacy in the treatment of VRE BSIs. Previous literature has mainly comprised case studies or series that have evaluated clinical outcomes with daptomycin in the treatment of VRE BSIs. Gallagher et al7 reported the results of a retrospective case series of 30 patients with VRE bacteremia who were treated with daptomycin. In this study, microbiologic cure was achieved in 80% of patients, with clinical success in 59% of the patients. In 2009, Mave et al4 compared clinical outcomes between daptomycin and linezolid in the treatment of VRE bacteremia. Reported results demonstrated a microbiologic cure rate of 90% for daptomycin versus 88% for linezolid.4 Moreover, there were no differences in mortality between the groups in our study. In 2010, Crank et al18 reported no differences in mortality (in‐hospital) for hospitalized patients with VRE BSIs treated with linezolid or daptomycin. Our results seem to be consistent with what has been published previously concerning clinical outcomes associated with linezolid or daptomycin in the treatment of VRE BSI.

The average daptomycin dose received in our patients was 6.1 mg/kg with doses ranging from 3.410.4 mg/kg. The underdosing as well as higher MICs to daptomycin may have contributed to a higher rate of recurrence. Previous reports state that Enterococcus species may have higher MICs to daptomycin than Staphylococcus or Streptococcus species; consequently, higher doses may be needed to adequately treat enterococcal infections.7 In the aforementioned study by Gallagher et al,7 doses of daptomycin 6 mg/kg were associated with a positive clinical outcome in 81% of patients compared with 31% if the dose used was <6 mg/kg. Linezolid is dosed 600 mg every 12 hours by mouth or intravenously, with no variations. There have been no studies comparing the uniform dosing of linezolid to the weight‐based dosing of daptomycin and their effects on outcomes.

Patients particularly susceptible to VRE infections include those with neutropenia and/or cancer, patients receiving long‐term hemodialysis, and liver transplant recipients.3, 22, 23 Upon review of this immunocompromised population, we noted no statistically significant differences in overall outcomes. A study by Kraft et al.24 supports the findings in our study that both drugs appear useful in the treatment of VRE bacteremia in patients with hematologic malignancy. We did identify a difference, albeit nonsignificant, in LOS for daptomycin versus linezolid in patients with a history of liver transplantation. Again, the level of care that these patients needed compared with the general population may explain this difference. As mentioned previously, another pertinent factor would be the dose of daptomycin used in these patients, because the dose can affect clinical success. Because all of the other patients had a similar LOS, we cannot determine that the increased LOS seen in liver transplant patients treated with daptomycin was solely due to daptomycin use. The reason for the increased LOS seems to be multifactorial. In the neutropenic population, a difference in LOS was also recognized, but follow‐up complete blood count values were not collected for these patients to determine whether linezolid contributed to further bone marrow suppression leading to an increase in LOS. For both of these patient populations, the number of patients included is very small (n = 21 for neutropenia total, n = 13 for liver transplant total), which can lead to a high degree of variance.

This study has several limitations. This was a retrospective review; therefore, we had no control over the selection of therapy. This may be reflected in an apparent preferential use of daptomycin in immunocompromised patients. Furthermore, 62% of linezolid patients and 54% of daptomycin patients received an antibiotic before initial therapy that could have potentially altered response rates. Due to the paucity of documentation surrounding initial site of infection, some of the positive cultures may represent potential contamination, because VRE may contaminate skin.25 Contamination seems implausible, however, because patients were seen by an infectious disease physician and had at least 1 documented positive VRE blood culture. We chose arbitrary definitions for clinical cure, microbiologic cure, microbiologic failure, recurrence, and reinfection. Previous studies have used their own definitions leading to discrepancies in reporting. Another limitation was that follow‐up cultures were not obtained on all of the patients, which was needed to determine rates of recurrence, reinfection, and microbiologic cure. MICs to daptomycin were not reported in 30% of our patients, potentially altering the recurrence rate seen in the daptomycin‐treated patients. Because clinical cure was not documented in the chart, it was inferred from the laboratory values and vital sign information. One investigator analyzed all of the values and made the determination of clinical cure, allowing for a consistent approach to data review.

In the face of the imposing threat of a highly resistant organism such as VRE with a limited number of efficacious antibiotics, antimicrobial selection becomes increasingly important and is requisite to clinical and microbiological success. To our knowledge, this is one of the largest studies to date comparing the efficacy of linezolid with that of daptomycin in the treatment of VRE bacteremia. Both of these agents are effective for the treatment of VRE BSIs. Nevertheless, specific factors related to the medication (eg, dose, route of administration) as well as the patient (eg, comorbid conditions, acuity of illness) should be taken into consideration when selecting an initial antimicrobial agent. Because the treatment of VRE BSIs continues to be a challenge, larger prospective randomized controlled trials are needed to corroborate our results and determine the optimal therapy for this serious infection.