Case 5-2005 — A 53-Year-Old Man with Depression and Sudden Shortness of Breath
http://www.100md.com
《新英格兰医药杂志》
Presentation of Case
Dr. Daniel Egan (Emergency Medicine, Massachusetts General Hospital and Brigham and Women's Hospital): A 53-year-old man came to the emergency department because of symptoms of depression.
The patient had been in good health until nine days earlier, when he was fired from his job. Since then, he had had a decrease in his energy level and appetite, was unable to concentrate, and had had difficulty sleeping. He had been spending all of his time in bed, where his wife had been feeding and bathing him. Because of the severity and persistence of his symptoms, his family brought him to the emergency department of this hospital for psychiatric evaluation.
The patient described his mood as sad, but he reported no thoughts of suicide. He stated that he had never felt like this in the past. He did not have fever, chills, nausea, vomiting, diarrhea, chest pain, or shortness of breath. He had had intermittent epigastric pain without radiation and a mild, nonproductive cough. His wife reported that he had had occasional night sweats that did not soak the bed.
At 35 years of age, he had been hospitalized for a problem with his heart, but he was unsure of what the diagnosis had been. He had had no operations. He took no medications and had no allergies. He smoked one pack of cigarettes every three days. He drank "a couple" of beers each week and had been hospitalized in the past for alcoholism. He did not use intravenous or recreational drugs. He lived with his wife and two children, who were well. He had not traveled recently or been exposed to animals.
The temperature was 36.7°C, the heart rate 102 beats per minute, and the blood pressure 97/52 mm Hg. The respiratory rate was 16 breaths per minute with an oxygen saturation of 98 percent while the patient was breathing ambient air. On physical examination, he was in no distress, but he was poorly groomed and malodorous, with a blunted affect. The general physical and neurologic examinations disclosed no abnormalities.
An electrocardiogram showed a normal sinus rhythm with first-degree atrioventricular block (PR interval, 358 msec) and left atrial enlargement (Figure 1A). There was no evidence of previous infarction or acute infarction. Blood samples were obtained for laboratory tests and a chest radiograph was ordered. Intravenous normal saline was administered.
Figure 1. Electrocardiograms.
The electrocardiogram of the patient obtained on admission shows a first-degree atrioventricular block and evidence of left atrial enlargement (Panel A). Five hours later, there is a new left bundle-branch block (Panel B).
While awaiting the results of the laboratory tests, the patient was seen by a member of the acute psychiatric service, who obtained the additional history that he had had auditory hallucinations, in which he heard two voices having a conversation. The patient met criteria for a major depressive disorder and admission to the psychiatric service was planned, pending medical clearance.
The chest radiograph revealed a right middle-lobe infiltrate (Figure 2A); intravenous ceftriaxone and oral azithromycin were administered. The results of laboratory tests are shown in Table 1. Because of the laboratory findings that showed leukocytosis and anemia, a decision was made to admit the patient to the medical service.
Figure 2. Chest Radiographs.
The initial radiograph reveals asymmetric airspace disease, a small right pleural effusion, and pulmonary vascular redistribution (Panel A). In Panel B, a radiograph of the chest obtained five hours later with portable equipment reveals diffuse pulmonary edema. An endotracheal tube and nasogastric tube are in place.
Table 1. Laboratory Values.
While awaiting an inpatient bed, the patient suddenly became short of breath and diaphoretic and reported pain in his chest. The oxygen saturation was 83 percent while he was receiving supplemental oxygen by way of a nonrebreather face mask. Physical examination revealed distended neck veins and rales to the apexes of both lungs. Cardiac examination revealed no murmur. The trachea was promptly intubated, and mechanical ventilation was begun. Electrocardiography showed a new left bundle-branch block and intermittent third-degree atrioventricular block (Figure 1B). A radiographic image of the chest obtained with portable equipment revealed pulmonary edema (Figure 2B). Packed red cells were transfused and furosemide and nitroglycerin were administered. Monitoring of the patient's cardiac rhythm revealed an intermittent third-degree heart block. External pacer pads were applied but not used because the episodes of heart block were brief.
Bedside transthoracic ultrasonographic examination showed aortic regurgitation. Transesophageal examination (Figure 3) showed an aortic-root abscess, a large area consistent with vegetation on the right aortic cusp and a leaflet fenestration (with moderate-to-severe aortic regurgitation), an area of vegetation on the base of the anterior mitral-valve leaflet, and mild mitral regurgitation. Left ventricular systolic function was mildly impaired. There was a small pericardial effusion.
Figure 3. Transesophageal Echocardiogram.
Panel A and Video Clip 1 of the Supplementary Appendix, available with the full text of this article at www.nejm.org, show a large area of vegetation (VEG, arrow in Panel A) on the anterior mitral valve (AMV), seen on the midesophageal four-chamber view. PMV denotes posterior mitral valve, and TV tricuspid valve. Panel B and Video Clip 2 of the Supplementary Appendix show a perforation (arrow in Panel B) of the right aortic valve (AV) leaflet and an abscess (triangles in Panel B) of the anterior aortic (Ao) root, seen on the midesophageal short-axis view of the aortic root. Panel C and Video Clip 3 of the Supplementary Appendix show a color Doppler study of a posteriorly directed jet of aortic regurgitation (AR, arrow in Panel C) against the anterior mitral leaflet. AoR denotes aortic root, and LVOT left ventricular outflow tract. A pulsed Doppler study (Panel D) shows the pan-diastolic flow reversal (arrows) in the distal ascending aorta. The scale to the right indicates flow velocity measured in centimeters per second, the signal above the baseline indicates forward flow, and the signal below the baseline indicates reverse flow.
A transvenous pacemaker was inserted, and the patient was admitted to the intensive care unit. Nafcillin, vancomycin, and gentamicin were administered. The next morning he was taken to the operating room. The aortic valve had been largely destroyed, with pus emanating from the intraventricular septum. There were fenestrations in both the right and noncoronary cusps. Purulent material extended laterally from the intraventricular septum, past the commissure of the right and left coronary cusps onto the anterior leaflet of the mitral valve, where there was another abscess, 1 cm in diameter. This was resected along with the majority of the anterior leaflet of the mitral valve. A homograft was used to replace the aortic valve, the left ventricular outflow tract, and the anterior leaflet of the mitral valve. A coronary-artery bypass graft was performed. The results of cultures of blood, abscess contents, and the resected tissue were negative.
The patient's postoperative course was complicated by acute renal and hepatic failure and candida sepsis, and he died after one month of hospitalization. Permission for an autopsy was declined.
Differential Diagnosis
Dr. Jo-Anne O. Shepard: The initial chest radiograph (Figure 2A) reveals asymmetric airspace disease, worse on the right side than on the left, a small right pleural effusion, and pulmonary vascular redistribution consistent with pulmonary venous hypertension. The heart and mediastinum are normal. The findings are suggestive of congestive heart failure with asymmetric pulmonary edema, but they could represent a right-sided pneumonia. A chest radiograph obtained with portable equipment after intubation (Figure 2B) shows the development of diffuse pulmonary edema.
Dr. Dali Fan: The transesophageal echocardiogram shows a multilobed area of vegetation, 2 cm by 1 cm by 0.75 cm, on the base of the anterior mitral leaflet (Figure 3A). There is mild mitral regurgitation. The aortic root is thickened, suggesting an aortic-root abscess, and an area of echolucency on the right coronary leaflet indicates perforation (Figure 3B). On color Doppler imaging, there is a jet of aortic regurgitation directed posteriorly across the left ventricular outflow tract, striking the anterior leaflet of the mitral valve (Figure 3C). The aortic-valve regurgitation causes pan-diastolic flow reversal on spectral Doppler imaging of the distal ascending aorta (Figure 3D).
Dr. Paul D. Biddinger: This case presents two common challenges that we face in the emergency department: the request for medical clearance of a patient with psychiatric symptoms and the need simultaneously to treat and determine the cause of acute, severe shortness of breath.
Medical Clearance of Psychiatric Patients
This patient and his family sought medical attention because of symptoms of depression. A diagnosis of a major depressive disorder was made, and medical clearance was requested before admitting him to the psychiatric service. "Medical clearance" is a term used in emergency departments for the medical evaluation of patients who are to be admitted with psychiatric symptoms. It is a request with two goals: to rule out organic disease as a cause of the patient's behavioral changes and to screen for an acute illness that might be more than an inpatient psychiatric facility can handle, whether or not the illness is the cause of the behavioral changes. Unfortunately, both the term and the process of medical clearance are plagued by conflict and confusion.
The term "medically cleared" is regarded by some as an oversimplification of a patient's health status and as an indication of poor communication among clinicians.1,2 It has been suggested that this label may impede further medical care, by implying that the patient has no physical health problems. One study reported that only 39 percent of psychiatric residents planned to perform a physical examination on patients once they had been stamped "medically cleared" by other clinicians.3 This assumption can be dangerous, since as many as 50 percent of psychiatric patients have coexisting illnesses that require ongoing evaluation and treatment.4,5
The other problem with the term is that the evaluations performed for medical clearance may be inconsistent from one institution or practitioner to another, inadequate, or both. In one study,2 nearly 33 percent of patients undergoing medical clearance had no vital signs recorded. In another,6 20 percent had no mental-status examination documented.
What should a medical-clearance evaluation consist of? Careful questioning to detect metabolic, neurologic, endocrine-related, or infectious causes of behavioral change is necessary, as is a thorough review of all medications and any history of illicit substance use. Some clues in the patient's history can help distinguish between organic (or medical) and functional (or psychiatric) causes of alterations in behavior (Table 2). Age is an important consideration: very young patients and patients over 40 years of age typically do not present to the emergency department with new onset of psychiatric symptoms. Although depression is a plausible explanation for some of this patient's symptoms, the presence of epigastric pain, cough, and night sweats, together with his older age, are warning signs of an organic process.
Table 2. Comparison of Causes of Alterations in Behavior.
Another important variable is the quality of the patient's change in mental status. Patients with medical illness may present with a sudden alteration in their behavior, and the mental status may follow a fluctuating course. In contrast, patients with psychiatric disease typically do not have altered levels of consciousness or disorientation.
Finally, a careful physical examination must be performed; abnormal vital signs or findings on physical examination should lead to consideration of organic causes of disease. This patient's physical examination was initially normal.
The question of what, if any, laboratory and imaging studies should be ordered in a situation such as this one is controversial. Screening laboratory tests have a low yield7,8 for patients with a known psychiatric disorder and a normal history and physical examination. However, in patients without a history of psychiatric disease, the odds of identifying a metabolic cause are higher. A study of 100 alert adult patients with new psychiatric symptoms or behavioral changes and no obvious signs of intoxication found that 63 had organic causes of their symptoms.9 The history suggested the cause in 27, the physical examination in 5, laboratory abnormalities in 44, imaging studies in 8, and lumbar puncture in 3 patients.
The patient in the case under discussion had symptoms of a major depressive disorder but had no history of psychiatric illness. For this reason, a thorough evaluation was both indicated and performed and led to the first clues to his underlying illness: the abnormal results of chest radiography and the elevated white-cell count.
Acute Shortness of Breath
While he was still in the emergency department, the patient became suddenly and profoundly short of breath. This development presented a very different challenge from the evaluation for medical clearance. He now needed simultaneous diagnosis and management of this life-threatening symptom. He was hypoxemic despite therapy with supplemental oxygen, had markedly labored breathing, and required immediate intubation. Thus, no further history could be obtained from the patient, and we had to rely on a rapid reassessment of the physical examination and on selected diagnostic tests to arrive at a diagnosis as quickly as possible.
The patient had rales to the apexes of both lungs, distended neck veins, and a chest radiograph that confirmed the signs of pulmonary edema. In retrospect, the infiltrate in the right middle lobe, seen on the initial radiograph, was believed to represent asymmetric pulmonary edema, rather than pneumonia. The electrocardiogram showed a new left bundle-branch block and intermittent third-degree heart block. Our differential diagnosis included acute myocardial infarction, a pulmonary embolus or pericardial effusion with tamponade, and valvular dysfunction. We needed a diagnostic method that would quickly help us distinguish among these. We elected not to have the patient undergo cardiac catheterization because of the rapidity of his decompensation, and instead we obtained an echocardiogram. The transesophageal echocardiogram then established the diagnosis.
Dr. Eric M. Isselbacher: This case presents two problems for the cardiologist: the differential diagnosis of acute (or so-called flash) pulmonary edema and the management of invasive endocarditis. Flash pulmonary edema is distinguished by rapid onset — appearing over a period of minutes — as we saw in this patient. Most cases of pulmonary edema that are described as acute are really subacute — arising over hours or even days. Recognizing this distinction can help to narrow the list of possible causes. In this case of flash pulmonary edema, there were three possible cardiac causes: left ventricular dysfunction, valvular dysfunction, and arrhythmia.
Acute left ventricular dysfunction can be either systolic or diastolic. The most common mechanism of acute left ventricular systolic dysfunction is acute myocardial infarction or ischemia that involves a large territory of myocardium, causing global left ventricular dysfunction. In this patient, there was no antecedent chest pain to suggest ischemia or infarction; instead, the chest pain appeared with the onset of his dyspnea. Acute myocarditis, a common cause of subacute pulmonary edema, can occasionally cause flash pulmonary edema, as can the apical ballooning syndrome. In cases of the apical ballooning syndrome, there is acute dilatation and akinesis of the midventricular and apical left ventricular walls, resulting in markedly impaired systolic function. The syndrome appears to be precipitated by acute emotional or physiological stress, by way of unknown mechanisms10; an echocardiogram is required to make the diagnosis. In this patient, the initial physical examination and electrocardiogram showed no evidence of myocarditis or valvular disease.
Diastolic dysfunction typically occurs in patients with long-standing hypertension and left ventricular hypertrophy and is often precipitated by acute or accelerated hypertension.11 However, since this patient had neither a history of hypertension nor hypertension on admission, diastolic dysfunction would be unlikely.
In patients with mitral or aortic stenosis, acute pulmonary edema can be precipitated by either volume overload or an arrhythmia. This patient did not have a history of valvular stenosis, and a murmur was not heard on the initial physical examination. A large left atrial myxoma can produce acute mitral stenosis and pulmonary edema, although this is rare.
Valvular regurgitation is more common than stenosis as a cause of flash pulmonary edema. In a patient with preexisting mitral-valve prolapse, mitral regurgitation can worsen abruptly when there is spontaneous rupture of the chordae tendineae. Apical tethering of the papillary muscles that results in incomplete mitral-valve closure and severe mitral regurgitation may develop in patients with left ventricular ischemia.12 Several days into the evolution of an acute myocardial infarction, rupture of the body or head of a papillary muscle can occur, resulting in acute severe mitral regurgitation.13 This patient had no history of mitral-valve prolapse and no evidence of an evolving myocardial infarction. Acute severe aortic regurgitation can also result from a proximal (type A) aortic dissection, and this can cause chest pain that occurs simultaneously with the onset of dyspnea, as in this patient.
Tachyarrhythmias may precipitate pulmonary edema in the setting of underlying left ventricular dysfunction or valvular dysfunction — particularly valvular stenosis — but rarely cause pulmonary edema in a patient with a structurally normal heart. Patients with mitral stenosis are particularly prone to atrial fibrillation, which can precipitate flash pulmonary edema. In this patient, there was no history of preexisting cardiac disease, and there was no evidence of a tachyarrhythmia before the onset of the dyspnea.
Finally, infective endocarditis can cause flash pulmonary edema. Mitral-valve endocarditis may cause regurgitation due to rupture of the chordae tendineae, abnormal leaflet coaptation, leaflet destruction producing perforation or tears, or invasion of mitral-annular tissue that produces a fistula between the left atrium and ventricle. Aortic-valve endocarditis may cause regurgitation due to cusp prolapse, valve destruction, or the formation of a fistula between the aorta and the left ventricular outflow tract or other cardiac chambers. This patient did not have fevers or chills, but he did have night sweats, a cough, anorexia, and leukocytosis; thus, although his symptoms are far from classic for endocarditis, they are suggestive enough that endocarditis should be considered. Even after the occurrence of flash pulmonary edema, no murmur was detected. Nevertheless, it is important to recognize that when examining an ill patient with acute respiratory distress in a noisy emergency department, it may be difficult to hear any but the loudest murmur. Moreover, the murmurs associated with acute severe aortic and mitral regurgitation are often much more subtle than one might expect. Thus, in this setting, the absence of a murmur does not rule out the presence of severe regurgitation. Finally, the pattern of progressive conduction abnormalities on the electrocardiogram — presentation with marked first-degree atrioventricular block, development of a left bundle-branch block, and then complete heart block — is consistent with rapidly progressive invasive endocarditis.
The echocardiogram that was obtained when the patient's breathing suddenly became labored showed normal left ventricular systolic function with no sign of myocardial infarction or the apical ballooning syndrome. A jet of aortic regurgitation was present, and although it did not appear initially to be severe, in the context of flash pulmonary edema one should suspect that the jet is in fact larger than it appears. Indeed, the transesophageal echocardiogram confirmed that the aortic regurgitation was severe, and it showed that the regurgitation was the result of endocarditis and not aortic dissection.
Discussion of Management
This is not a typical case of invasive infective endocarditis. "Invasive" refers to extension of the infection from the valve into the adjacent cardiac tissues, which occurs in 13 to 30 percent of cases of endocarditis.14 Invasive endocarditis involves the aortic valve more often than the mitral valve,15 but it may involve both, as it did in this patient. The possibility of invasive endocarditis should be considered when fever persists or recurs despite appropriate antibiotic therapy, or when new or evolving cardiac-conduction abnormalities appear,16 as they did in this patient.
First-degree atrioventricular block is the most common cardiac-conduction abnormality, but only a minority of cases progress to complete heart block. About 50 percent of cases of invasive endocarditis are caused by Staphylococcus aureus.17 Streptococcus viridans species and Staphylococcus epidermidis are also relatively common pathogens, particularly on prosthetic heart valves. In some cases, even with invasive disease, no organism is identified — this is so-called culture-negative endocarditis. Surgical specimens may also fail to yield the culprit organism, as was the case here.
As soon as the diagnosis of endocarditis is suspected, blood cultures should be performed, and antibiotic therapy administered, as was done with this patient. Surgery to repair the damaged cardiac structures and to try to eradicate the infection is the treatment of choice unless there is a clear contraindication. When invasive endocarditis involves the aortic valve, an aortic-root homograft — cryopreserved human tissue, which is resistant to infection — is typically used. A homograft can be used that includes the anterior leaflet of the mitral valve, which permits the surgeon to extend a repair into the anterior mitral leaflet should it be involved by the endocarditis, as it was in this patient, avoiding replacement of the mitral valve.
Dr. Nancy Lee Harris (Pathology): At the time of the patient's initial visit to the emergency room, should the first-degree heart block have been a clue that he had a cardiac problem?
Dr. Isselbacher: Although first-degree atrioventricular block is common, this patient had marked first-degree atrioventricular block, which is more unusual. I think that if there had been any signs or symptoms of endocarditis, the staff of the emergency department would have recognized the importance of the atrioventricular block immediately.
Dr. Harris: Dr. Parks, would you like to comment?
Dr. Larry Parks (Acute Psychiatry): The patient presented with typical features of a major depression with psychotic features. There was a precipitant or source of stress in the form of his job loss that led us to suspect a primary psychiatric diagnosis. The features that suggested the need for a comprehensive medical evaluation were the epigastric pain, the night sweats, and the cough.
Dr. Egan: After the patient died, a report from the Centers for Disease Control and Prevention noted that serologic tests had been positive for bartonella (Bartonella henselae IgG, 1:128; B. quintana IgG, 1:512). Could bartonella have caused this case of endocarditis?
Dr. Isselbacher: Bartonella has been recognized with increasing frequency over the past decade as a cause of culture-negative endocarditis18 and may cause as many as 3 percent of all cases of endocarditis. It can be difficult to culture from blood or tissue. It is usually identified on serologic testing, so all patients with suspected culture-negative endocarditis should be tested. The polymerase chain reaction (PCR) can be used to amplify the bartonella DNA from surgical specimens. Typically, bartonella causes endocarditis in patients who are homeless, who have alcoholism, or who have been exposed to cats.19 This patient was not homeless and did not have a cat, although he did have a history of alcoholism. About 15 percent of patients with bartonella-associated endocarditis are afebrile. Bartonella has a propensity to infect the aortic valve, often a valve that had been normal, although it is typically not invasive. Although some features fit with this patient's presentation and others do not, it seems probable that bartonella was the cause of the endocarditis.
Dr. Harris: Unfortunately, all the resected tissue was submitted for culture, so no tissue is available for PCR analysis for microorganisms.
Final Diagnosis
Invasive endocarditis, possibly due to bartonella.
Source Information
From the Department of Emergency Medicine (P.D.B.), the Cardiology Division, Department of Medicine (E.M.I., D.F.), and the Department of Radiology (J.O.S.), Massachusetts General Hospital; and the Departments of Surgery (P.D.B.), Medicine (E.M.I., D.F.), and Radiology (J.O.S.), Harvard Medical School.
References
Weissberg MP. Emergency room medical clearance: an educational problem. Am J Psychiatry 1979;136:787-790.
Tintinalli JE, Peacock FW IV, Wright MA. Emergency medical evaluation of psychiatric patients. Ann Emerg Med 1994;23:859-862.
Greden JF, Casariego JI. Controversies in psychiatric education: a survey of residents' attitudes. Am J Psychiatry 1975;132:270-274.
Carlson RJ, Nayar N, Suh M. Physical disorders among emergency psychiatric patients. Can J Psychiatry 1981;26:65-67.
Summers WK, Muniz RA, Read MR, Marsh GM. The psychiatric physical examination. II. Findings in 75 unselected psychiatric patients. J Clin Psychiatry 1981;42:99-102.
Riba M, Hale M. Medical clearance: fact or fiction in the hospital emergency room. Psychosomatics 1990;31:400-404.
Olshaker JS, Browne B, Jerrard DA, Prendergast H, Stair TO. Medical clearance and screening of psychiatric patients in the emergency department. Acad Emerg Med 1997;4:124-128.
Korn CS, Currier GW, Henderson DO. "Medical clearance" of psychiatric patients without medical complaints in the emergency department. J Emerg Med 2000;18:173-176.
Henneman PL, Mendoza R, Lewis RJ. Prospective evaluation of emergency department medical clearance. Ann Emerg Med 1994;24:672-677.
Abe Y, Kondo M, Matsuoka R, Araki M, Dohyama K, Tanio H. Assessment of clinical features in transient left ventricular apical ballooning. J Am Coll Cardiol 2003;41:737-742.
Kramer K, Kirkman P, Kitzman D, Little WC. Flash pulmonary edema: association with hypertension and reoccurrence despite coronary revascularization. Am Heart J 2000;140:451-455.
Pierard LA, Lancellotti P. The role of ischemic mitral regurgitation in the pathogenesis of acute pulmonary edema. N Engl J Med 2004;351:1627-1634.
Case Records of the Massachusetts General Hospital (Case 29-2002). N Engl J Med 2002;347:921-928.
Shafran SD. Infective endocarditis and perivalvular abscess: a dangerous duo. CMAJ 2002;167:38-39.
Chan KL. Early clinical course and long-term outcome of patients with infective endocarditis complicated by perivalvular abscess. CMAJ 2002;167:19-24.
Brecker SJD, Pepper JR, Eykyn SJ. Aortic root abscess. Heart 1999;82:260-262.
DiNubile MJ, Calderwood SB, Steinhaus DM, Karchmer AW. Cardiac conduction abnormalities complicating native valve active infective endocarditis. Am J Cardiol 1986;58:1213-1217.
Raoult D, Fournier PE, Drancourt M, et al. Diagnosis of 22 new cases of Bartonella endocarditis. Ann Intern Med 1996;125:646-652.
Drancourt M, Mainardi JL, Brouqui P, et al. Bartonella (Rochalimaea) quintana endocarditis in three homeless men. N Engl J Med 1995;332:419-423.(Paul D. Biddinger, M.D., )
Dr. Daniel Egan (Emergency Medicine, Massachusetts General Hospital and Brigham and Women's Hospital): A 53-year-old man came to the emergency department because of symptoms of depression.
The patient had been in good health until nine days earlier, when he was fired from his job. Since then, he had had a decrease in his energy level and appetite, was unable to concentrate, and had had difficulty sleeping. He had been spending all of his time in bed, where his wife had been feeding and bathing him. Because of the severity and persistence of his symptoms, his family brought him to the emergency department of this hospital for psychiatric evaluation.
The patient described his mood as sad, but he reported no thoughts of suicide. He stated that he had never felt like this in the past. He did not have fever, chills, nausea, vomiting, diarrhea, chest pain, or shortness of breath. He had had intermittent epigastric pain without radiation and a mild, nonproductive cough. His wife reported that he had had occasional night sweats that did not soak the bed.
At 35 years of age, he had been hospitalized for a problem with his heart, but he was unsure of what the diagnosis had been. He had had no operations. He took no medications and had no allergies. He smoked one pack of cigarettes every three days. He drank "a couple" of beers each week and had been hospitalized in the past for alcoholism. He did not use intravenous or recreational drugs. He lived with his wife and two children, who were well. He had not traveled recently or been exposed to animals.
The temperature was 36.7°C, the heart rate 102 beats per minute, and the blood pressure 97/52 mm Hg. The respiratory rate was 16 breaths per minute with an oxygen saturation of 98 percent while the patient was breathing ambient air. On physical examination, he was in no distress, but he was poorly groomed and malodorous, with a blunted affect. The general physical and neurologic examinations disclosed no abnormalities.
An electrocardiogram showed a normal sinus rhythm with first-degree atrioventricular block (PR interval, 358 msec) and left atrial enlargement (Figure 1A). There was no evidence of previous infarction or acute infarction. Blood samples were obtained for laboratory tests and a chest radiograph was ordered. Intravenous normal saline was administered.
Figure 1. Electrocardiograms.
The electrocardiogram of the patient obtained on admission shows a first-degree atrioventricular block and evidence of left atrial enlargement (Panel A). Five hours later, there is a new left bundle-branch block (Panel B).
While awaiting the results of the laboratory tests, the patient was seen by a member of the acute psychiatric service, who obtained the additional history that he had had auditory hallucinations, in which he heard two voices having a conversation. The patient met criteria for a major depressive disorder and admission to the psychiatric service was planned, pending medical clearance.
The chest radiograph revealed a right middle-lobe infiltrate (Figure 2A); intravenous ceftriaxone and oral azithromycin were administered. The results of laboratory tests are shown in Table 1. Because of the laboratory findings that showed leukocytosis and anemia, a decision was made to admit the patient to the medical service.
Figure 2. Chest Radiographs.
The initial radiograph reveals asymmetric airspace disease, a small right pleural effusion, and pulmonary vascular redistribution (Panel A). In Panel B, a radiograph of the chest obtained five hours later with portable equipment reveals diffuse pulmonary edema. An endotracheal tube and nasogastric tube are in place.
Table 1. Laboratory Values.
While awaiting an inpatient bed, the patient suddenly became short of breath and diaphoretic and reported pain in his chest. The oxygen saturation was 83 percent while he was receiving supplemental oxygen by way of a nonrebreather face mask. Physical examination revealed distended neck veins and rales to the apexes of both lungs. Cardiac examination revealed no murmur. The trachea was promptly intubated, and mechanical ventilation was begun. Electrocardiography showed a new left bundle-branch block and intermittent third-degree atrioventricular block (Figure 1B). A radiographic image of the chest obtained with portable equipment revealed pulmonary edema (Figure 2B). Packed red cells were transfused and furosemide and nitroglycerin were administered. Monitoring of the patient's cardiac rhythm revealed an intermittent third-degree heart block. External pacer pads were applied but not used because the episodes of heart block were brief.
Bedside transthoracic ultrasonographic examination showed aortic regurgitation. Transesophageal examination (Figure 3) showed an aortic-root abscess, a large area consistent with vegetation on the right aortic cusp and a leaflet fenestration (with moderate-to-severe aortic regurgitation), an area of vegetation on the base of the anterior mitral-valve leaflet, and mild mitral regurgitation. Left ventricular systolic function was mildly impaired. There was a small pericardial effusion.
Figure 3. Transesophageal Echocardiogram.
Panel A and Video Clip 1 of the Supplementary Appendix, available with the full text of this article at www.nejm.org, show a large area of vegetation (VEG, arrow in Panel A) on the anterior mitral valve (AMV), seen on the midesophageal four-chamber view. PMV denotes posterior mitral valve, and TV tricuspid valve. Panel B and Video Clip 2 of the Supplementary Appendix show a perforation (arrow in Panel B) of the right aortic valve (AV) leaflet and an abscess (triangles in Panel B) of the anterior aortic (Ao) root, seen on the midesophageal short-axis view of the aortic root. Panel C and Video Clip 3 of the Supplementary Appendix show a color Doppler study of a posteriorly directed jet of aortic regurgitation (AR, arrow in Panel C) against the anterior mitral leaflet. AoR denotes aortic root, and LVOT left ventricular outflow tract. A pulsed Doppler study (Panel D) shows the pan-diastolic flow reversal (arrows) in the distal ascending aorta. The scale to the right indicates flow velocity measured in centimeters per second, the signal above the baseline indicates forward flow, and the signal below the baseline indicates reverse flow.
A transvenous pacemaker was inserted, and the patient was admitted to the intensive care unit. Nafcillin, vancomycin, and gentamicin were administered. The next morning he was taken to the operating room. The aortic valve had been largely destroyed, with pus emanating from the intraventricular septum. There were fenestrations in both the right and noncoronary cusps. Purulent material extended laterally from the intraventricular septum, past the commissure of the right and left coronary cusps onto the anterior leaflet of the mitral valve, where there was another abscess, 1 cm in diameter. This was resected along with the majority of the anterior leaflet of the mitral valve. A homograft was used to replace the aortic valve, the left ventricular outflow tract, and the anterior leaflet of the mitral valve. A coronary-artery bypass graft was performed. The results of cultures of blood, abscess contents, and the resected tissue were negative.
The patient's postoperative course was complicated by acute renal and hepatic failure and candida sepsis, and he died after one month of hospitalization. Permission for an autopsy was declined.
Differential Diagnosis
Dr. Jo-Anne O. Shepard: The initial chest radiograph (Figure 2A) reveals asymmetric airspace disease, worse on the right side than on the left, a small right pleural effusion, and pulmonary vascular redistribution consistent with pulmonary venous hypertension. The heart and mediastinum are normal. The findings are suggestive of congestive heart failure with asymmetric pulmonary edema, but they could represent a right-sided pneumonia. A chest radiograph obtained with portable equipment after intubation (Figure 2B) shows the development of diffuse pulmonary edema.
Dr. Dali Fan: The transesophageal echocardiogram shows a multilobed area of vegetation, 2 cm by 1 cm by 0.75 cm, on the base of the anterior mitral leaflet (Figure 3A). There is mild mitral regurgitation. The aortic root is thickened, suggesting an aortic-root abscess, and an area of echolucency on the right coronary leaflet indicates perforation (Figure 3B). On color Doppler imaging, there is a jet of aortic regurgitation directed posteriorly across the left ventricular outflow tract, striking the anterior leaflet of the mitral valve (Figure 3C). The aortic-valve regurgitation causes pan-diastolic flow reversal on spectral Doppler imaging of the distal ascending aorta (Figure 3D).
Dr. Paul D. Biddinger: This case presents two common challenges that we face in the emergency department: the request for medical clearance of a patient with psychiatric symptoms and the need simultaneously to treat and determine the cause of acute, severe shortness of breath.
Medical Clearance of Psychiatric Patients
This patient and his family sought medical attention because of symptoms of depression. A diagnosis of a major depressive disorder was made, and medical clearance was requested before admitting him to the psychiatric service. "Medical clearance" is a term used in emergency departments for the medical evaluation of patients who are to be admitted with psychiatric symptoms. It is a request with two goals: to rule out organic disease as a cause of the patient's behavioral changes and to screen for an acute illness that might be more than an inpatient psychiatric facility can handle, whether or not the illness is the cause of the behavioral changes. Unfortunately, both the term and the process of medical clearance are plagued by conflict and confusion.
The term "medically cleared" is regarded by some as an oversimplification of a patient's health status and as an indication of poor communication among clinicians.1,2 It has been suggested that this label may impede further medical care, by implying that the patient has no physical health problems. One study reported that only 39 percent of psychiatric residents planned to perform a physical examination on patients once they had been stamped "medically cleared" by other clinicians.3 This assumption can be dangerous, since as many as 50 percent of psychiatric patients have coexisting illnesses that require ongoing evaluation and treatment.4,5
The other problem with the term is that the evaluations performed for medical clearance may be inconsistent from one institution or practitioner to another, inadequate, or both. In one study,2 nearly 33 percent of patients undergoing medical clearance had no vital signs recorded. In another,6 20 percent had no mental-status examination documented.
What should a medical-clearance evaluation consist of? Careful questioning to detect metabolic, neurologic, endocrine-related, or infectious causes of behavioral change is necessary, as is a thorough review of all medications and any history of illicit substance use. Some clues in the patient's history can help distinguish between organic (or medical) and functional (or psychiatric) causes of alterations in behavior (Table 2). Age is an important consideration: very young patients and patients over 40 years of age typically do not present to the emergency department with new onset of psychiatric symptoms. Although depression is a plausible explanation for some of this patient's symptoms, the presence of epigastric pain, cough, and night sweats, together with his older age, are warning signs of an organic process.
Table 2. Comparison of Causes of Alterations in Behavior.
Another important variable is the quality of the patient's change in mental status. Patients with medical illness may present with a sudden alteration in their behavior, and the mental status may follow a fluctuating course. In contrast, patients with psychiatric disease typically do not have altered levels of consciousness or disorientation.
Finally, a careful physical examination must be performed; abnormal vital signs or findings on physical examination should lead to consideration of organic causes of disease. This patient's physical examination was initially normal.
The question of what, if any, laboratory and imaging studies should be ordered in a situation such as this one is controversial. Screening laboratory tests have a low yield7,8 for patients with a known psychiatric disorder and a normal history and physical examination. However, in patients without a history of psychiatric disease, the odds of identifying a metabolic cause are higher. A study of 100 alert adult patients with new psychiatric symptoms or behavioral changes and no obvious signs of intoxication found that 63 had organic causes of their symptoms.9 The history suggested the cause in 27, the physical examination in 5, laboratory abnormalities in 44, imaging studies in 8, and lumbar puncture in 3 patients.
The patient in the case under discussion had symptoms of a major depressive disorder but had no history of psychiatric illness. For this reason, a thorough evaluation was both indicated and performed and led to the first clues to his underlying illness: the abnormal results of chest radiography and the elevated white-cell count.
Acute Shortness of Breath
While he was still in the emergency department, the patient became suddenly and profoundly short of breath. This development presented a very different challenge from the evaluation for medical clearance. He now needed simultaneous diagnosis and management of this life-threatening symptom. He was hypoxemic despite therapy with supplemental oxygen, had markedly labored breathing, and required immediate intubation. Thus, no further history could be obtained from the patient, and we had to rely on a rapid reassessment of the physical examination and on selected diagnostic tests to arrive at a diagnosis as quickly as possible.
The patient had rales to the apexes of both lungs, distended neck veins, and a chest radiograph that confirmed the signs of pulmonary edema. In retrospect, the infiltrate in the right middle lobe, seen on the initial radiograph, was believed to represent asymmetric pulmonary edema, rather than pneumonia. The electrocardiogram showed a new left bundle-branch block and intermittent third-degree heart block. Our differential diagnosis included acute myocardial infarction, a pulmonary embolus or pericardial effusion with tamponade, and valvular dysfunction. We needed a diagnostic method that would quickly help us distinguish among these. We elected not to have the patient undergo cardiac catheterization because of the rapidity of his decompensation, and instead we obtained an echocardiogram. The transesophageal echocardiogram then established the diagnosis.
Dr. Eric M. Isselbacher: This case presents two problems for the cardiologist: the differential diagnosis of acute (or so-called flash) pulmonary edema and the management of invasive endocarditis. Flash pulmonary edema is distinguished by rapid onset — appearing over a period of minutes — as we saw in this patient. Most cases of pulmonary edema that are described as acute are really subacute — arising over hours or even days. Recognizing this distinction can help to narrow the list of possible causes. In this case of flash pulmonary edema, there were three possible cardiac causes: left ventricular dysfunction, valvular dysfunction, and arrhythmia.
Acute left ventricular dysfunction can be either systolic or diastolic. The most common mechanism of acute left ventricular systolic dysfunction is acute myocardial infarction or ischemia that involves a large territory of myocardium, causing global left ventricular dysfunction. In this patient, there was no antecedent chest pain to suggest ischemia or infarction; instead, the chest pain appeared with the onset of his dyspnea. Acute myocarditis, a common cause of subacute pulmonary edema, can occasionally cause flash pulmonary edema, as can the apical ballooning syndrome. In cases of the apical ballooning syndrome, there is acute dilatation and akinesis of the midventricular and apical left ventricular walls, resulting in markedly impaired systolic function. The syndrome appears to be precipitated by acute emotional or physiological stress, by way of unknown mechanisms10; an echocardiogram is required to make the diagnosis. In this patient, the initial physical examination and electrocardiogram showed no evidence of myocarditis or valvular disease.
Diastolic dysfunction typically occurs in patients with long-standing hypertension and left ventricular hypertrophy and is often precipitated by acute or accelerated hypertension.11 However, since this patient had neither a history of hypertension nor hypertension on admission, diastolic dysfunction would be unlikely.
In patients with mitral or aortic stenosis, acute pulmonary edema can be precipitated by either volume overload or an arrhythmia. This patient did not have a history of valvular stenosis, and a murmur was not heard on the initial physical examination. A large left atrial myxoma can produce acute mitral stenosis and pulmonary edema, although this is rare.
Valvular regurgitation is more common than stenosis as a cause of flash pulmonary edema. In a patient with preexisting mitral-valve prolapse, mitral regurgitation can worsen abruptly when there is spontaneous rupture of the chordae tendineae. Apical tethering of the papillary muscles that results in incomplete mitral-valve closure and severe mitral regurgitation may develop in patients with left ventricular ischemia.12 Several days into the evolution of an acute myocardial infarction, rupture of the body or head of a papillary muscle can occur, resulting in acute severe mitral regurgitation.13 This patient had no history of mitral-valve prolapse and no evidence of an evolving myocardial infarction. Acute severe aortic regurgitation can also result from a proximal (type A) aortic dissection, and this can cause chest pain that occurs simultaneously with the onset of dyspnea, as in this patient.
Tachyarrhythmias may precipitate pulmonary edema in the setting of underlying left ventricular dysfunction or valvular dysfunction — particularly valvular stenosis — but rarely cause pulmonary edema in a patient with a structurally normal heart. Patients with mitral stenosis are particularly prone to atrial fibrillation, which can precipitate flash pulmonary edema. In this patient, there was no history of preexisting cardiac disease, and there was no evidence of a tachyarrhythmia before the onset of the dyspnea.
Finally, infective endocarditis can cause flash pulmonary edema. Mitral-valve endocarditis may cause regurgitation due to rupture of the chordae tendineae, abnormal leaflet coaptation, leaflet destruction producing perforation or tears, or invasion of mitral-annular tissue that produces a fistula between the left atrium and ventricle. Aortic-valve endocarditis may cause regurgitation due to cusp prolapse, valve destruction, or the formation of a fistula between the aorta and the left ventricular outflow tract or other cardiac chambers. This patient did not have fevers or chills, but he did have night sweats, a cough, anorexia, and leukocytosis; thus, although his symptoms are far from classic for endocarditis, they are suggestive enough that endocarditis should be considered. Even after the occurrence of flash pulmonary edema, no murmur was detected. Nevertheless, it is important to recognize that when examining an ill patient with acute respiratory distress in a noisy emergency department, it may be difficult to hear any but the loudest murmur. Moreover, the murmurs associated with acute severe aortic and mitral regurgitation are often much more subtle than one might expect. Thus, in this setting, the absence of a murmur does not rule out the presence of severe regurgitation. Finally, the pattern of progressive conduction abnormalities on the electrocardiogram — presentation with marked first-degree atrioventricular block, development of a left bundle-branch block, and then complete heart block — is consistent with rapidly progressive invasive endocarditis.
The echocardiogram that was obtained when the patient's breathing suddenly became labored showed normal left ventricular systolic function with no sign of myocardial infarction or the apical ballooning syndrome. A jet of aortic regurgitation was present, and although it did not appear initially to be severe, in the context of flash pulmonary edema one should suspect that the jet is in fact larger than it appears. Indeed, the transesophageal echocardiogram confirmed that the aortic regurgitation was severe, and it showed that the regurgitation was the result of endocarditis and not aortic dissection.
Discussion of Management
This is not a typical case of invasive infective endocarditis. "Invasive" refers to extension of the infection from the valve into the adjacent cardiac tissues, which occurs in 13 to 30 percent of cases of endocarditis.14 Invasive endocarditis involves the aortic valve more often than the mitral valve,15 but it may involve both, as it did in this patient. The possibility of invasive endocarditis should be considered when fever persists or recurs despite appropriate antibiotic therapy, or when new or evolving cardiac-conduction abnormalities appear,16 as they did in this patient.
First-degree atrioventricular block is the most common cardiac-conduction abnormality, but only a minority of cases progress to complete heart block. About 50 percent of cases of invasive endocarditis are caused by Staphylococcus aureus.17 Streptococcus viridans species and Staphylococcus epidermidis are also relatively common pathogens, particularly on prosthetic heart valves. In some cases, even with invasive disease, no organism is identified — this is so-called culture-negative endocarditis. Surgical specimens may also fail to yield the culprit organism, as was the case here.
As soon as the diagnosis of endocarditis is suspected, blood cultures should be performed, and antibiotic therapy administered, as was done with this patient. Surgery to repair the damaged cardiac structures and to try to eradicate the infection is the treatment of choice unless there is a clear contraindication. When invasive endocarditis involves the aortic valve, an aortic-root homograft — cryopreserved human tissue, which is resistant to infection — is typically used. A homograft can be used that includes the anterior leaflet of the mitral valve, which permits the surgeon to extend a repair into the anterior mitral leaflet should it be involved by the endocarditis, as it was in this patient, avoiding replacement of the mitral valve.
Dr. Nancy Lee Harris (Pathology): At the time of the patient's initial visit to the emergency room, should the first-degree heart block have been a clue that he had a cardiac problem?
Dr. Isselbacher: Although first-degree atrioventricular block is common, this patient had marked first-degree atrioventricular block, which is more unusual. I think that if there had been any signs or symptoms of endocarditis, the staff of the emergency department would have recognized the importance of the atrioventricular block immediately.
Dr. Harris: Dr. Parks, would you like to comment?
Dr. Larry Parks (Acute Psychiatry): The patient presented with typical features of a major depression with psychotic features. There was a precipitant or source of stress in the form of his job loss that led us to suspect a primary psychiatric diagnosis. The features that suggested the need for a comprehensive medical evaluation were the epigastric pain, the night sweats, and the cough.
Dr. Egan: After the patient died, a report from the Centers for Disease Control and Prevention noted that serologic tests had been positive for bartonella (Bartonella henselae IgG, 1:128; B. quintana IgG, 1:512). Could bartonella have caused this case of endocarditis?
Dr. Isselbacher: Bartonella has been recognized with increasing frequency over the past decade as a cause of culture-negative endocarditis18 and may cause as many as 3 percent of all cases of endocarditis. It can be difficult to culture from blood or tissue. It is usually identified on serologic testing, so all patients with suspected culture-negative endocarditis should be tested. The polymerase chain reaction (PCR) can be used to amplify the bartonella DNA from surgical specimens. Typically, bartonella causes endocarditis in patients who are homeless, who have alcoholism, or who have been exposed to cats.19 This patient was not homeless and did not have a cat, although he did have a history of alcoholism. About 15 percent of patients with bartonella-associated endocarditis are afebrile. Bartonella has a propensity to infect the aortic valve, often a valve that had been normal, although it is typically not invasive. Although some features fit with this patient's presentation and others do not, it seems probable that bartonella was the cause of the endocarditis.
Dr. Harris: Unfortunately, all the resected tissue was submitted for culture, so no tissue is available for PCR analysis for microorganisms.
Final Diagnosis
Invasive endocarditis, possibly due to bartonella.
Source Information
From the Department of Emergency Medicine (P.D.B.), the Cardiology Division, Department of Medicine (E.M.I., D.F.), and the Department of Radiology (J.O.S.), Massachusetts General Hospital; and the Departments of Surgery (P.D.B.), Medicine (E.M.I., D.F.), and Radiology (J.O.S.), Harvard Medical School.
References
Weissberg MP. Emergency room medical clearance: an educational problem. Am J Psychiatry 1979;136:787-790.
Tintinalli JE, Peacock FW IV, Wright MA. Emergency medical evaluation of psychiatric patients. Ann Emerg Med 1994;23:859-862.
Greden JF, Casariego JI. Controversies in psychiatric education: a survey of residents' attitudes. Am J Psychiatry 1975;132:270-274.
Carlson RJ, Nayar N, Suh M. Physical disorders among emergency psychiatric patients. Can J Psychiatry 1981;26:65-67.
Summers WK, Muniz RA, Read MR, Marsh GM. The psychiatric physical examination. II. Findings in 75 unselected psychiatric patients. J Clin Psychiatry 1981;42:99-102.
Riba M, Hale M. Medical clearance: fact or fiction in the hospital emergency room. Psychosomatics 1990;31:400-404.
Olshaker JS, Browne B, Jerrard DA, Prendergast H, Stair TO. Medical clearance and screening of psychiatric patients in the emergency department. Acad Emerg Med 1997;4:124-128.
Korn CS, Currier GW, Henderson DO. "Medical clearance" of psychiatric patients without medical complaints in the emergency department. J Emerg Med 2000;18:173-176.
Henneman PL, Mendoza R, Lewis RJ. Prospective evaluation of emergency department medical clearance. Ann Emerg Med 1994;24:672-677.
Abe Y, Kondo M, Matsuoka R, Araki M, Dohyama K, Tanio H. Assessment of clinical features in transient left ventricular apical ballooning. J Am Coll Cardiol 2003;41:737-742.
Kramer K, Kirkman P, Kitzman D, Little WC. Flash pulmonary edema: association with hypertension and reoccurrence despite coronary revascularization. Am Heart J 2000;140:451-455.
Pierard LA, Lancellotti P. The role of ischemic mitral regurgitation in the pathogenesis of acute pulmonary edema. N Engl J Med 2004;351:1627-1634.
Case Records of the Massachusetts General Hospital (Case 29-2002). N Engl J Med 2002;347:921-928.
Shafran SD. Infective endocarditis and perivalvular abscess: a dangerous duo. CMAJ 2002;167:38-39.
Chan KL. Early clinical course and long-term outcome of patients with infective endocarditis complicated by perivalvular abscess. CMAJ 2002;167:19-24.
Brecker SJD, Pepper JR, Eykyn SJ. Aortic root abscess. Heart 1999;82:260-262.
DiNubile MJ, Calderwood SB, Steinhaus DM, Karchmer AW. Cardiac conduction abnormalities complicating native valve active infective endocarditis. Am J Cardiol 1986;58:1213-1217.
Raoult D, Fournier PE, Drancourt M, et al. Diagnosis of 22 new cases of Bartonella endocarditis. Ann Intern Med 1996;125:646-652.
Drancourt M, Mainardi JL, Brouqui P, et al. Bartonella (Rochalimaea) quintana endocarditis in three homeless men. N Engl J Med 1995;332:419-423.(Paul D. Biddinger, M.D., )