Don't Know Much about History
http://www.100md.com
《新英格兰医药杂志》
A 20-year-old man who had immigrated to the United States from Mexico three months earlier presented to the emergency department, reporting that he had had weight loss and pain in his hip and back on the right side. The patient did not speak English; the initial history was obtained with the use of the limited Spanish of team members and with the help of untrained interpreters. Two weeks before presentation, the patient slipped and fell while washing floors. Although he did not recall any pain immediately after the event, pain in the right side of his lower back and hip developed several days later. The pain was described as intermittent and stabbing, and it increased with movement. He noted that he had had decreased appetite for several weeks and had unintentionally lost 16 kg. He also said he had had occasional bitemporal headache, dizziness, and nausea and was easily fatigued and mildly short of breath. He said he did not have fever, chills, night sweats, confusion, diarrhea, constipation, vomiting, urinary urgency, cough, or easy bruising.
Although back pain is frequently benign, several factors — most important, this patient's weight loss — make me worry about serious causes, in particular infection or cancer. One obvious concern is spinal epidural abscess, which might be caused by occult bacteremia, tuberculosis, or a fungal infection, such as coccidioidomycosis. Musculoskeletal pain can be a sign of infective endocarditis. I am also considering amebic liver abscess, since the patient is a Mexican man with pain on the right side of his body. Lymphoma, leukemia, or germ-cell tumors may at times occur with or come to medical attention because of back pain, and they are among the most common cancers at his age. I want to know the patient's travel history, whether he has any risk factors for infection with the human immunodeficiency virus (HIV), and whether he has been exposed to others who have been ill. A physical examination is needed, with particular attention to fever, lymphadenopathy, or masses. In addition, I want to know if the "hip pain" is truly coming from the hip.
The patient had noted a loss of appetite since discontinuing valproic acid three weeks before presentation. He had been taking this medication since he was 13 years of age, after he had had a grand mal seizure. At that time, he was hospitalized for three days in Mexico and was told that he would need to take antiseizure medications for the rest of his life. The patient said that he did not use alcohol, tobacco, or illicit drugs. He had had one female sexual partner. He was living with a cousin and working in a grocery store.
I am puzzled about his association of loss of appetite with the discontinuation of his valproic acid, and I wonder why he stopped taking the medication. The history of seizures in a person originally from Mexico makes it likely that he has neurocysticercosis, which, in areas where the disease is endemic, is a more common cause of seizures than idiopathic epilepsy. It is tempting to apply Occam's razor in an attempt to explain all of this young man's problems with one disease. Neurocysticercosis may involve the spinal cord as well as the brain, and it is possible that he is having radicular pain from a subarachnoid cyst.
The patient was alert but appeared cachectic. His height was 165.1 cm and his weight 47 kg. His temperature was 37.3°C, blood pressure 95/63 mm Hg, pulse 93 beats per minute, and respiratory rate 12 breaths per minute. The results of his lung and cardiac examinations showed no abnormalities. He had palpable splenomegaly and an enlarged liver that spanned 12 cm, but he did not have abdominal tenderness. He did have mild tenderness with percussion of his right lower back and with external rotation of the right hip. A neurologic examination revealed a slow, unsteady gait, global weakness without focal deficits, and normal, symmetric reflexes. The result of a Romberg's test was negative, and there was no tremor or dysmetria. No rash, petechiae, lymphadenopathy, or scleral icterus was noted. Funduscopic and genitourinary examinations were not performed. The remainder of the patient's physical examination revealed no abnormalities.
His chronically ill and wasted appearance cannot be accounted for by a two-week illness; his physical presentation increases my suspicion of cancer or chronic infection. In addition, adrenal insufficiency should be considered, especially since this finding may occur with disseminated tuberculosis or metastatic cancer. The presence of splenomegaly leads me to consider either splenic congestion, usually from portal hypertension due to liver disease, or an infiltrative disease. Diffuse involvement of the liver by cancer (in particular, small-cell lung cancer) can lead to portal hypertension. A lymphoma might cause hepatosplenomegaly without portal hypertension. The absence of peripheral lymphadenopathy makes lymphoma a less likely cause but by no means rules it out. Testing for HIV infection is warranted.
Tuberculosis and a fungal infection remain serious considerations. Extracranial cysticercosis is improbable, since the liver and spleen are not common sites of involvement. Although the absence of fever and heart murmur also makes endocarditis improbable, I would not rule it out in this cachectic man with splenomegaly and back pain. One normal temperature does not rule out intermittent fever, and subtle heart murmurs such as that produced by aortic insufficiency may be easily overlooked.
The patient had a white-cell count of 4000 per cubic milliliter, with 52 percent polymorphonuclear cells, 4 percent band forms, 4 percent monocytes, 39 percent lymphocytes, and 1 percent metamyelocytes. His hemoglobin level was 13.8 g per deciliter, with evidence of teardrop cells, acanthocytes, poikilocytes, and ovalocytes on a peripheral-blood smear. The platelet count was 20,000 per cubic milliliter. The results of a complete chemistry panel showed no abnormal values except for the following: alkaline phosphatase, 199 U per liter (normal range, 40 to 120 U per liter); aspartate transaminase, 147 U per liter (normal range, 16 to 50 U per liter); alanine transaminase, 75 U per liter (normal range, 12 to 61 U per liter); and lactate dehydrogenase, 1115 U per liter (normal range, 313 to 618 U per liter). Coagulation studies were normal. The amylase level was 136 U per liter (normal range, 30 to 110 U per liter), and the lipase level was 419 U per liter (normal range, 23 to 300 U per liter). The level of valproic acid was undetectable, and the results of a toxicologic screen were negative. The erythrocyte sedimentation rate was normal at 4 mm per hour. The C-reactive protein level was 3.1 mg per deciliter (normal range, less than 0.8 mg per deciliter). Radiographs of the patient's chest, right hip, and lumbar spine showed no abnormalities. A computed tomographic (CT) scan of the head showed a calcified lesion, 2 mm in length, in the right frontal lobe. A CT scan of the abdomen and pelvis revealed diffuse hepatic edema and splenomegaly. Blood and urine cultures were obtained.
The presence of misshapen red cells and thrombocytopenia suggests bone marrow invasion, perhaps due to a disseminated infection or tumor. These findings would not be expected in infective endocarditis. The combination of thrombocytopenia and fragmented red cells may also be seen in microangiopathic states such as the hemolytic–uremic syndrome or malignant hypertension, but none of these fit the clinical context. It is possible that the patient's enlarged spleen is sequestering platelets, but this would not cause damaged red cells to appear in the peripheral blood. Tumor is less probable than infection; no masses were found on imaging studies. Miliary tuberculosis could be responsible for all the clinical findings and should be ruled out. The normal chest radiograph does not rule out this diagnosis, since a substantial proportion of patients with miliary tuberculosis present with a normal chest radiograph. Disseminated tuberculosis and endocarditis are generally associated with a high erythrocyte sedimentation rate; however, a normal erythrocyte sedimentation rate does not rule out these infections. I suggest that a bone marrow biopsy or liver biopsy be performed. I would start the patient on empirical treatment for tuberculosis while awaiting the results of these studies.
The patient was treated with ibuprofen and acetaminophen for his pain. In the hospital, he was noted to have nightly fevers ranging from 38.5°C to 40.5°C. He had occasional chills, but he said that he never felt febrile. On hospital day 2, a new systolic heart murmur was noted, and he began to report pain in his right elbow, similar to that in his right hip. There was no initial growth reported from either set of blood cultures obtained at admission. The test for HIV was negative. A tuberculin skin test was negative, but the control tests for candida and for mumps were also negative. The monospot test for mononucleosis and the acute hepatitis panel for hepatitis A, B, and C were all negative.
Given the finding of fever, it is possible that he has been febrile throughout his illness. Tuberculosis remains my most urgent concern. A negative tuberculin skin test is not helpful in the context of negative controls.
A bone marrow aspirate showed a normocellular marrow with mild myeloid hyperplasia, decreased iron stores, and rare noncaseating granulomas (Figure 1). A bone marrow smear was negative for acid-fast bacilli. A bone scan obtained with use of technetium-99–labeled methylene diphosphonate scintigraphy revealed mild, asymmetric uptake in the patient's right ankle and right sacroiliac joint, which was believed to be due to degenerative changes; however, osteomyelitis could not be ruled out. Transthoracic and transesophageal echocardiograms showed no evidence of vegetations; the heart function was normal.
Figure 1. Bone Marrow Aspirate Showing a Noncaseating Granuloma (Arrow) (Hematoxylin and Eosin).
The granulomas found in the bone marrow make it highly probable that the patient has a disseminated granulomatous disease, and tuberculosis is the most likely candidate. It is also possible that he has a systemic fungal infection such as coccidioidomycosis. Other causes of noncaseating granulomas, such as sarcoidosis or Hodgkin's disease, are less probable.
During hospitalization, additional details of the history were obtained with the help of an interpreter. The patient reported that he had worked for three years in a slaughterhouse in Mexico, where he had had daily exposure to carcasses of swine, cattle, and sheep.
This is a critical piece of historical information. If the case had been reframed as that of a "patient with classic fever of unknown origin," then any history of exposure to animals would have been explicitly sought. He might have been a typical abattoir worker of Mexico and, therefore, regularly exposed to Brucella abortus and B. suis — both of which are endemic in Mexican livestock. His illness, including the hepatosplenomegaly and bone marrow findings, is consistent with brucellosis.
Given this additional history, and the patient's ill appearance, I would empirically treat for brucellosis by adding doxycycline to his current treatment regimen. I would continue empirical treatment for both this and tuberculosis while waiting for the results of serologic tests for brucella and for cultures.
On the seventh hospital day, serologic testing for brucella was positive, with a titer greater than 1:160. The next day, the repeated blood cultures and bone marrow culture were positive for gram-negative coccobacilli; the organisms were subsequently identified as B. abortus. The patient was treated with doxycycline and rifampin. He became afebrile, his heart murmur disappeared, and the pain in his hip, back, and elbow diminished within a few days.
As is often noted, if the diagnosis is not suspected on the basis of the history, it will probably not be made. Taking a thorough history from most patients requires time and patience. Obtaining a history through an interpreter requires extra time and patience. But by doing so, the clinicians caring for this patient were able to make the correct diagnosis.
Commentary
Brucellosis is a zoonotic disease that can involve any organ or system of the body. Most cases in humans are caused by contact with infected animals or animal products, such as unpasteurized milk and cheese.1 Four species are responsible for most human cases: B. abortus (found in cattle), B. melitensis (found in sheep and goats), B. suis (found in swine), and B. canis (found in dogs). The clinical presentation can be acute or progressive, usually starting with nonspecific symptoms, such as fever, sweats, weight loss, arthralgias, and myalgias. In one series of patients infected with B. melitensis, the most common findings on physical examination included a diffusely enlarged liver or spleen (present in 38 percent and 22 percent, respectively), osteoarticular abnormalities, such as spondylitis and sacroiliitis (present in 21 percent), relative bradycardia (present in 21 percent), and lymphadenopathy (present in 9 percent).2 Both cultures and serologic tests are useful for the diagnosis of brucellosis. In the series mentioned above, when cultures were performed under ideal circumstances (defined as an absence of antibiotic treatment in the preceding 72 hours and incubation for at least four weeks), approximately 70 percent were positive for brucella. Recently, higher rates of positive cultures have been reported (91 percent in acute brucellosis and 74 percent in chronic brucellosis).3
Although brucellosis remains common in developing countries, the disease is rare in the United States, with about 100 cases reported annually. More than half of these cases have occurred in Hispanic patients, according to the category checked on the survey forms.4 The incubation period is usually between 7 days and 3 months, although incubation periods as long as 10 months have been reported.5 A thorough travel history as well as a history of exposure to animals and exotic foods are usually critical to making the diagnosis.1
In hindsight, this patient had many classic symptoms of brucellosis; however, it took several days before the clinical team caring for the patient made the correct diagnosis. The initial presentation was compatible with other, more common diseases that the discussant considered, including tuberculosis, endocarditis, fungal infections, hematologic cancers, and an HIV-related illness. The treating clinicians' initial assessment was similar, and their initial evaluation was directed toward a diagnosis based on consideration of these diseases. As the discussant correctly noted, despite a negative tuberculin skin test, the patient could still have had miliary tuberculosis,6 which can be rapidly fatal if untreated.
After admission, the patient was noted to have nightly fevers. The discussant suggested reframing the problem as fever of unknown origin, although there was no documentation that the patient had had three weeks of fever (as required by the formal criteria7 for this entity). Nevertheless, the patient did meet most criteria for a comprehensive evaluation for fever of unknown origin, including a physical examination, basic laboratory tests, blood cultures, chest radiography, and abdominal imaging.8 A critical part of the evaluation also includes a thorough history, which was initially compromised by the language barrier. Fortunately, the use of a professional interpreter eventually allowed the clinical team to discover crucial information — the patient's occupational exposure to the carcasses of swine, cattle, and sheep — thereby alerting the clinicians to the strong possibility of exposure to brucella. After this exposure history became known, serologic studies for brucella were obtained and monitoring of blood cultures was extended, leading to the correct diagnosis.
More than 21 million people in the United States, or 8.1 percent of the population, reported to the 2000 Census Bureau that they spoke English less than very well or not at all, as compared with 6.1 percent of the population in the 1990 U.S. Census.9 Furthermore, new regulations requiring adequate access to interpreters for patients covered by Medicare will make the need for interpreters more acute.10 The use of family members or untrained Spanish-speaking employees is not ideal. Indeed, the use of trained interpreters has been shown to increase access to care,11 and it may also improve patients' understanding of their disease. Ensuring access to adequate interpreters remains a challenge, but as our case demonstrates, it may be critical to obtaining an accurate history.
It is well recognized that, in the majority of cases, the diagnosis can be made after history taking alone.12 A study performed three decades ago to determine the relative value of the history, physical examination, and laboratory testing in making diagnoses found that the correct diagnosis was determined after the completion of only the history in 82 percent of patients.13 A similar study performed in 1992 yielded similar results, with the history leading to the correct diagnosis in 76 percent of the cases.14 Given the increasing number of non–English-speaking patients in the United States, the use of skilled interpreters to assist with the illness history is more important than ever.
Dr. Saint is supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant from the Agency for Healthcare Research and Quality (P20-HS11540).
We are indebted to Merle Sande, M.D., for assistance in the preparation of the manuscript, and to Bashar Dabbas, M.D., for his assistance with the microphotography.
Source Information
From the Department of Medicine, University of Utah, Salt Lake City (S.D., C.T.); the Department of Veterans Affairs Health Services Research and Development Center of Excellence and the Department of Medicine, University of Michigan — both in Ann Arbor (S.S.); and the Department of Medicine, Legacy Good Samaritan and Emanuel Hospitals and the Department of Medicine, Oregon Health and Sciences University — both in Portland (S.R.J.).
Address reprint requests to Dr. Dames at the University of Utah Medical Center, Division of Rheumatology 4B200, School of Medicine, 30 N. 1900 E, Salt Lake City, UT 84132, or at shelby.dames@hsc.utah.edu.
References
Young EJ. An overview of human brucellosis. Clin Infect Dis 1995;21:283-290.
Colmenero JD, Reguera JM, Martos F, et al. Complications associated with Brucella melitensis infection: a study of 530 cases. Medicine (Baltimore) 1996;75:195-211.
Mantur BG, Mangalgi SS. Evaluation of conventional Castaneda and lysis centrifugation blood culture techniques for diagnosis of human brucellosis. J Clin Microbiol 2004;42:4327-4328.
Chang MH, Glynn MK, Groseclose SL. Endemic, notifiable bioterrorism-related diseases, United States, 1992-1999. Emerg Infect Dis 2003;9:556-564.
Georghiou PR, Young EJ. Prolonged incubation in brucellosis. Lancet 1991;337:1543-1543.
Kim JH, Langston AA, Gallis HA. Miliary tuberculosis: epidemiology, clinical manifestations, diagnosis, and outcome. Rev Infect Dis 1990;12:583-590.
Petersdorf RG, Beeson PB. Fever of unexplained origin: report on 100 cases. Medicine (Baltimore) 1961;40:1-30.
Knockaert DC, Vanderschueren S, Blockmans D. Fever of unknown origin in adults: 40 years on. J Intern Med 2003;253:263-275.
Census Bureau. Language use and English-speaking ability: 2000. Census 2000 Brief. (Accessed May 9, 2005, at http://www.census.gov/prod/2003pubs/c2kbr-29.pdf.)
Guidance to federal financial assistance recipients regarding Title VI Prohibition Against National Origin Discrimination Affecting Limited English Proficient Persons. Washington, D.C.: Department of Health and Human Services, 2003. (Accessed May 9, 2005, at http://www.hhs.gov/ocr/lep/revisedlep.html.)
Jacobs EA, Lauderdale DS, Meltzer D, Shorey JM, Levinson W, Thisted RA. Impact of interpreter services on delivery of health care to limited-English-proficient patients. J Gen Intern Med 2001;16:468-474.
Platt R. Two essays on the practice of medicine. Manch Univ Med Sch Gaz 1947;27:139-45.
Hampton JR, Harrison MJG, Mitchell JRA, Prichard JS, Seymour C. Relative contributions of history-taking, physical examination, and laboratory investigation to diagnosis and management of medical outpatients. Br Med J 1975;2:486-489.
Peterson MC, Holbrook JH, Von Hales D, Smith NL, Staker LV. Contributions of the history, physical examination, and laboratory investigation in making medical diagnoses. West J Med 1992;156:163-165.(Shelby Dames, M.D., Claud)
Although back pain is frequently benign, several factors — most important, this patient's weight loss — make me worry about serious causes, in particular infection or cancer. One obvious concern is spinal epidural abscess, which might be caused by occult bacteremia, tuberculosis, or a fungal infection, such as coccidioidomycosis. Musculoskeletal pain can be a sign of infective endocarditis. I am also considering amebic liver abscess, since the patient is a Mexican man with pain on the right side of his body. Lymphoma, leukemia, or germ-cell tumors may at times occur with or come to medical attention because of back pain, and they are among the most common cancers at his age. I want to know the patient's travel history, whether he has any risk factors for infection with the human immunodeficiency virus (HIV), and whether he has been exposed to others who have been ill. A physical examination is needed, with particular attention to fever, lymphadenopathy, or masses. In addition, I want to know if the "hip pain" is truly coming from the hip.
The patient had noted a loss of appetite since discontinuing valproic acid three weeks before presentation. He had been taking this medication since he was 13 years of age, after he had had a grand mal seizure. At that time, he was hospitalized for three days in Mexico and was told that he would need to take antiseizure medications for the rest of his life. The patient said that he did not use alcohol, tobacco, or illicit drugs. He had had one female sexual partner. He was living with a cousin and working in a grocery store.
I am puzzled about his association of loss of appetite with the discontinuation of his valproic acid, and I wonder why he stopped taking the medication. The history of seizures in a person originally from Mexico makes it likely that he has neurocysticercosis, which, in areas where the disease is endemic, is a more common cause of seizures than idiopathic epilepsy. It is tempting to apply Occam's razor in an attempt to explain all of this young man's problems with one disease. Neurocysticercosis may involve the spinal cord as well as the brain, and it is possible that he is having radicular pain from a subarachnoid cyst.
The patient was alert but appeared cachectic. His height was 165.1 cm and his weight 47 kg. His temperature was 37.3°C, blood pressure 95/63 mm Hg, pulse 93 beats per minute, and respiratory rate 12 breaths per minute. The results of his lung and cardiac examinations showed no abnormalities. He had palpable splenomegaly and an enlarged liver that spanned 12 cm, but he did not have abdominal tenderness. He did have mild tenderness with percussion of his right lower back and with external rotation of the right hip. A neurologic examination revealed a slow, unsteady gait, global weakness without focal deficits, and normal, symmetric reflexes. The result of a Romberg's test was negative, and there was no tremor or dysmetria. No rash, petechiae, lymphadenopathy, or scleral icterus was noted. Funduscopic and genitourinary examinations were not performed. The remainder of the patient's physical examination revealed no abnormalities.
His chronically ill and wasted appearance cannot be accounted for by a two-week illness; his physical presentation increases my suspicion of cancer or chronic infection. In addition, adrenal insufficiency should be considered, especially since this finding may occur with disseminated tuberculosis or metastatic cancer. The presence of splenomegaly leads me to consider either splenic congestion, usually from portal hypertension due to liver disease, or an infiltrative disease. Diffuse involvement of the liver by cancer (in particular, small-cell lung cancer) can lead to portal hypertension. A lymphoma might cause hepatosplenomegaly without portal hypertension. The absence of peripheral lymphadenopathy makes lymphoma a less likely cause but by no means rules it out. Testing for HIV infection is warranted.
Tuberculosis and a fungal infection remain serious considerations. Extracranial cysticercosis is improbable, since the liver and spleen are not common sites of involvement. Although the absence of fever and heart murmur also makes endocarditis improbable, I would not rule it out in this cachectic man with splenomegaly and back pain. One normal temperature does not rule out intermittent fever, and subtle heart murmurs such as that produced by aortic insufficiency may be easily overlooked.
The patient had a white-cell count of 4000 per cubic milliliter, with 52 percent polymorphonuclear cells, 4 percent band forms, 4 percent monocytes, 39 percent lymphocytes, and 1 percent metamyelocytes. His hemoglobin level was 13.8 g per deciliter, with evidence of teardrop cells, acanthocytes, poikilocytes, and ovalocytes on a peripheral-blood smear. The platelet count was 20,000 per cubic milliliter. The results of a complete chemistry panel showed no abnormal values except for the following: alkaline phosphatase, 199 U per liter (normal range, 40 to 120 U per liter); aspartate transaminase, 147 U per liter (normal range, 16 to 50 U per liter); alanine transaminase, 75 U per liter (normal range, 12 to 61 U per liter); and lactate dehydrogenase, 1115 U per liter (normal range, 313 to 618 U per liter). Coagulation studies were normal. The amylase level was 136 U per liter (normal range, 30 to 110 U per liter), and the lipase level was 419 U per liter (normal range, 23 to 300 U per liter). The level of valproic acid was undetectable, and the results of a toxicologic screen were negative. The erythrocyte sedimentation rate was normal at 4 mm per hour. The C-reactive protein level was 3.1 mg per deciliter (normal range, less than 0.8 mg per deciliter). Radiographs of the patient's chest, right hip, and lumbar spine showed no abnormalities. A computed tomographic (CT) scan of the head showed a calcified lesion, 2 mm in length, in the right frontal lobe. A CT scan of the abdomen and pelvis revealed diffuse hepatic edema and splenomegaly. Blood and urine cultures were obtained.
The presence of misshapen red cells and thrombocytopenia suggests bone marrow invasion, perhaps due to a disseminated infection or tumor. These findings would not be expected in infective endocarditis. The combination of thrombocytopenia and fragmented red cells may also be seen in microangiopathic states such as the hemolytic–uremic syndrome or malignant hypertension, but none of these fit the clinical context. It is possible that the patient's enlarged spleen is sequestering platelets, but this would not cause damaged red cells to appear in the peripheral blood. Tumor is less probable than infection; no masses were found on imaging studies. Miliary tuberculosis could be responsible for all the clinical findings and should be ruled out. The normal chest radiograph does not rule out this diagnosis, since a substantial proportion of patients with miliary tuberculosis present with a normal chest radiograph. Disseminated tuberculosis and endocarditis are generally associated with a high erythrocyte sedimentation rate; however, a normal erythrocyte sedimentation rate does not rule out these infections. I suggest that a bone marrow biopsy or liver biopsy be performed. I would start the patient on empirical treatment for tuberculosis while awaiting the results of these studies.
The patient was treated with ibuprofen and acetaminophen for his pain. In the hospital, he was noted to have nightly fevers ranging from 38.5°C to 40.5°C. He had occasional chills, but he said that he never felt febrile. On hospital day 2, a new systolic heart murmur was noted, and he began to report pain in his right elbow, similar to that in his right hip. There was no initial growth reported from either set of blood cultures obtained at admission. The test for HIV was negative. A tuberculin skin test was negative, but the control tests for candida and for mumps were also negative. The monospot test for mononucleosis and the acute hepatitis panel for hepatitis A, B, and C were all negative.
Given the finding of fever, it is possible that he has been febrile throughout his illness. Tuberculosis remains my most urgent concern. A negative tuberculin skin test is not helpful in the context of negative controls.
A bone marrow aspirate showed a normocellular marrow with mild myeloid hyperplasia, decreased iron stores, and rare noncaseating granulomas (Figure 1). A bone marrow smear was negative for acid-fast bacilli. A bone scan obtained with use of technetium-99–labeled methylene diphosphonate scintigraphy revealed mild, asymmetric uptake in the patient's right ankle and right sacroiliac joint, which was believed to be due to degenerative changes; however, osteomyelitis could not be ruled out. Transthoracic and transesophageal echocardiograms showed no evidence of vegetations; the heart function was normal.
Figure 1. Bone Marrow Aspirate Showing a Noncaseating Granuloma (Arrow) (Hematoxylin and Eosin).
The granulomas found in the bone marrow make it highly probable that the patient has a disseminated granulomatous disease, and tuberculosis is the most likely candidate. It is also possible that he has a systemic fungal infection such as coccidioidomycosis. Other causes of noncaseating granulomas, such as sarcoidosis or Hodgkin's disease, are less probable.
During hospitalization, additional details of the history were obtained with the help of an interpreter. The patient reported that he had worked for three years in a slaughterhouse in Mexico, where he had had daily exposure to carcasses of swine, cattle, and sheep.
This is a critical piece of historical information. If the case had been reframed as that of a "patient with classic fever of unknown origin," then any history of exposure to animals would have been explicitly sought. He might have been a typical abattoir worker of Mexico and, therefore, regularly exposed to Brucella abortus and B. suis — both of which are endemic in Mexican livestock. His illness, including the hepatosplenomegaly and bone marrow findings, is consistent with brucellosis.
Given this additional history, and the patient's ill appearance, I would empirically treat for brucellosis by adding doxycycline to his current treatment regimen. I would continue empirical treatment for both this and tuberculosis while waiting for the results of serologic tests for brucella and for cultures.
On the seventh hospital day, serologic testing for brucella was positive, with a titer greater than 1:160. The next day, the repeated blood cultures and bone marrow culture were positive for gram-negative coccobacilli; the organisms were subsequently identified as B. abortus. The patient was treated with doxycycline and rifampin. He became afebrile, his heart murmur disappeared, and the pain in his hip, back, and elbow diminished within a few days.
As is often noted, if the diagnosis is not suspected on the basis of the history, it will probably not be made. Taking a thorough history from most patients requires time and patience. Obtaining a history through an interpreter requires extra time and patience. But by doing so, the clinicians caring for this patient were able to make the correct diagnosis.
Commentary
Brucellosis is a zoonotic disease that can involve any organ or system of the body. Most cases in humans are caused by contact with infected animals or animal products, such as unpasteurized milk and cheese.1 Four species are responsible for most human cases: B. abortus (found in cattle), B. melitensis (found in sheep and goats), B. suis (found in swine), and B. canis (found in dogs). The clinical presentation can be acute or progressive, usually starting with nonspecific symptoms, such as fever, sweats, weight loss, arthralgias, and myalgias. In one series of patients infected with B. melitensis, the most common findings on physical examination included a diffusely enlarged liver or spleen (present in 38 percent and 22 percent, respectively), osteoarticular abnormalities, such as spondylitis and sacroiliitis (present in 21 percent), relative bradycardia (present in 21 percent), and lymphadenopathy (present in 9 percent).2 Both cultures and serologic tests are useful for the diagnosis of brucellosis. In the series mentioned above, when cultures were performed under ideal circumstances (defined as an absence of antibiotic treatment in the preceding 72 hours and incubation for at least four weeks), approximately 70 percent were positive for brucella. Recently, higher rates of positive cultures have been reported (91 percent in acute brucellosis and 74 percent in chronic brucellosis).3
Although brucellosis remains common in developing countries, the disease is rare in the United States, with about 100 cases reported annually. More than half of these cases have occurred in Hispanic patients, according to the category checked on the survey forms.4 The incubation period is usually between 7 days and 3 months, although incubation periods as long as 10 months have been reported.5 A thorough travel history as well as a history of exposure to animals and exotic foods are usually critical to making the diagnosis.1
In hindsight, this patient had many classic symptoms of brucellosis; however, it took several days before the clinical team caring for the patient made the correct diagnosis. The initial presentation was compatible with other, more common diseases that the discussant considered, including tuberculosis, endocarditis, fungal infections, hematologic cancers, and an HIV-related illness. The treating clinicians' initial assessment was similar, and their initial evaluation was directed toward a diagnosis based on consideration of these diseases. As the discussant correctly noted, despite a negative tuberculin skin test, the patient could still have had miliary tuberculosis,6 which can be rapidly fatal if untreated.
After admission, the patient was noted to have nightly fevers. The discussant suggested reframing the problem as fever of unknown origin, although there was no documentation that the patient had had three weeks of fever (as required by the formal criteria7 for this entity). Nevertheless, the patient did meet most criteria for a comprehensive evaluation for fever of unknown origin, including a physical examination, basic laboratory tests, blood cultures, chest radiography, and abdominal imaging.8 A critical part of the evaluation also includes a thorough history, which was initially compromised by the language barrier. Fortunately, the use of a professional interpreter eventually allowed the clinical team to discover crucial information — the patient's occupational exposure to the carcasses of swine, cattle, and sheep — thereby alerting the clinicians to the strong possibility of exposure to brucella. After this exposure history became known, serologic studies for brucella were obtained and monitoring of blood cultures was extended, leading to the correct diagnosis.
More than 21 million people in the United States, or 8.1 percent of the population, reported to the 2000 Census Bureau that they spoke English less than very well or not at all, as compared with 6.1 percent of the population in the 1990 U.S. Census.9 Furthermore, new regulations requiring adequate access to interpreters for patients covered by Medicare will make the need for interpreters more acute.10 The use of family members or untrained Spanish-speaking employees is not ideal. Indeed, the use of trained interpreters has been shown to increase access to care,11 and it may also improve patients' understanding of their disease. Ensuring access to adequate interpreters remains a challenge, but as our case demonstrates, it may be critical to obtaining an accurate history.
It is well recognized that, in the majority of cases, the diagnosis can be made after history taking alone.12 A study performed three decades ago to determine the relative value of the history, physical examination, and laboratory testing in making diagnoses found that the correct diagnosis was determined after the completion of only the history in 82 percent of patients.13 A similar study performed in 1992 yielded similar results, with the history leading to the correct diagnosis in 76 percent of the cases.14 Given the increasing number of non–English-speaking patients in the United States, the use of skilled interpreters to assist with the illness history is more important than ever.
Dr. Saint is supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant from the Agency for Healthcare Research and Quality (P20-HS11540).
We are indebted to Merle Sande, M.D., for assistance in the preparation of the manuscript, and to Bashar Dabbas, M.D., for his assistance with the microphotography.
Source Information
From the Department of Medicine, University of Utah, Salt Lake City (S.D., C.T.); the Department of Veterans Affairs Health Services Research and Development Center of Excellence and the Department of Medicine, University of Michigan — both in Ann Arbor (S.S.); and the Department of Medicine, Legacy Good Samaritan and Emanuel Hospitals and the Department of Medicine, Oregon Health and Sciences University — both in Portland (S.R.J.).
Address reprint requests to Dr. Dames at the University of Utah Medical Center, Division of Rheumatology 4B200, School of Medicine, 30 N. 1900 E, Salt Lake City, UT 84132, or at shelby.dames@hsc.utah.edu.
References
Young EJ. An overview of human brucellosis. Clin Infect Dis 1995;21:283-290.
Colmenero JD, Reguera JM, Martos F, et al. Complications associated with Brucella melitensis infection: a study of 530 cases. Medicine (Baltimore) 1996;75:195-211.
Mantur BG, Mangalgi SS. Evaluation of conventional Castaneda and lysis centrifugation blood culture techniques for diagnosis of human brucellosis. J Clin Microbiol 2004;42:4327-4328.
Chang MH, Glynn MK, Groseclose SL. Endemic, notifiable bioterrorism-related diseases, United States, 1992-1999. Emerg Infect Dis 2003;9:556-564.
Georghiou PR, Young EJ. Prolonged incubation in brucellosis. Lancet 1991;337:1543-1543.
Kim JH, Langston AA, Gallis HA. Miliary tuberculosis: epidemiology, clinical manifestations, diagnosis, and outcome. Rev Infect Dis 1990;12:583-590.
Petersdorf RG, Beeson PB. Fever of unexplained origin: report on 100 cases. Medicine (Baltimore) 1961;40:1-30.
Knockaert DC, Vanderschueren S, Blockmans D. Fever of unknown origin in adults: 40 years on. J Intern Med 2003;253:263-275.
Census Bureau. Language use and English-speaking ability: 2000. Census 2000 Brief. (Accessed May 9, 2005, at http://www.census.gov/prod/2003pubs/c2kbr-29.pdf.)
Guidance to federal financial assistance recipients regarding Title VI Prohibition Against National Origin Discrimination Affecting Limited English Proficient Persons. Washington, D.C.: Department of Health and Human Services, 2003. (Accessed May 9, 2005, at http://www.hhs.gov/ocr/lep/revisedlep.html.)
Jacobs EA, Lauderdale DS, Meltzer D, Shorey JM, Levinson W, Thisted RA. Impact of interpreter services on delivery of health care to limited-English-proficient patients. J Gen Intern Med 2001;16:468-474.
Platt R. Two essays on the practice of medicine. Manch Univ Med Sch Gaz 1947;27:139-45.
Hampton JR, Harrison MJG, Mitchell JRA, Prichard JS, Seymour C. Relative contributions of history-taking, physical examination, and laboratory investigation to diagnosis and management of medical outpatients. Br Med J 1975;2:486-489.
Peterson MC, Holbrook JH, Von Hales D, Smith NL, Staker LV. Contributions of the history, physical examination, and laboratory investigation in making medical diagnoses. West J Med 1992;156:163-165.(Shelby Dames, M.D., Claud)