Bertha Johnson is back with pneumonia again. The ED doctor on the telephone sounded both matter‐of‐fact and mildly bored when I answered her page about another admission to the hospitalist service. I hadn't met Mrs. Johnson previously, but came to know her and Douglas, her only son, well over the next few days.

The Johnsons were facing a difficult choice. Bertha was now bedbound and quadriplegic following a 40‐year battle with multiple sclerosis and gradually mounting disability. She was cognitively intact and had a solid grasp of medical realities, but was hard of hearing and quite dysarthric. Forming even short phrases and sentences took great effort. However tenuous, this ability to speak allowed her to communicate with those she loved. She had been admitted thrice in the past year with aspiration pneumonia, as she was unable to clear her secretions reliably. Repeated bronchoscopies demonstrated an inability to protect her airway. Douglas, who was also her health care proxy, favored proceeding with the tracheostomy suggested by the pulmonary team. On a prior admission, he had been distressed when his mother refused this intervention. Now she was back with the identical problem and had been given the same recommendation by her doctors. It was particularly difficult for me to discuss these sensitive issues when I had not previously met either Bertha or her son. I spoke to her primary care physician over the phone and he agreed with the need for tracheostomy. The pulmonary team had been involved in discussions about tracheostomy in all of her hospital admissions, providing continuity of care in the process. Ultimately, it was my responsibility to help Bertha and Douglas come to a decision.

After multiple discussions between the Johnsons and me, a consensus emerged to proceed with the tracheostomy. We recognized that the procedure would increase her care needs and arranged for a stay in a skilled nursing facility to provide access to round‐the‐clock suctioning. The evening prior to the tracheostomy, the floor nurse and I reviewed the procedure to ensure that Bertha was fully prepared. What followed resulted in a drastic change of plan. Bertha emphasized that she did not want to lose her only means of communication, even if the surgery would prolong her life. She admitted that she reluctantly agreed to the procedure only to please her son and doctors, because they believed it to be in her best interest. Her fear of the prospect of death from drowning in her own secretions was much less than her fear of silence and isolation that would result from her loss of speech. I shared her misgivings with Douglas, and she admitted to him that she had only agreed to the procedure for his sake. We cancelled the surgery.

Douglas later revealed that he also had been ambivalent about the procedure for sometime, as it would necessitate a nursing home stay and the loss of the caretakers who had cared for his mother so wonderfully for many years. Bertha lived alone in her own home with the help of visiting nurses and patient care assistants Douglas paid for out‐of‐pocket. Douglas lived and worked in a city over a hundred miles away, but managed to visit several times a month to facilitate his mother's care. He had supported the procedure only because the doctors had said it was the only way to avoid future pneumonias. The idea of a tracheostomy was definitively abandoned once and for all.

The Johnsons wanted Bertha to return to her home, but hospital case managers felt this would be an unsafe plan of care as she was alone for several hours a day. She was largely immobile and unable to escape if there were a fire or other emergency. Also, the caretakers were not trained to use the suction equipment, and the visiting nurses would only be available intermittently. The home care staff felt they could no longer meet her needs and declined to resume her care. Douglas became very frustrated by the delays and protracted negotiations, enough so that he threatened to sue the institution for taking over my mother's life. The threat of litigation is usually a cry for help that reflects either miscommunication or the suffering of a conflicted family.

As their hospitalist, I hoped to advocate for both the patient and Douglas while coordinating the overall care plan. I had always received consistent responses from the Johnsons, but other staff members noted that Douglas had expressed shifting views on the best site of care for his mother. At Bertha's request, I convened a meeting with her, Douglas, the social worker, case manager, visiting and staff nurses, the palliative care nurse, and floor manager. Prior to this, I met with all involved health care providers to ensure we understood each other's abilities and limitations regarding Bertha's care. As I entered the room for the family meeting, I knew it was ultimately the patient's choice whether she wanted to return home or notas long as she understood the risks involved.

During the meeting, all the team members explained the dilemmas they faced in planning for a safe disposition. Douglas's response illuminated his devotion and love for his ailing mother. He had known all along that it would be less expensive and burdensome for him for Bertha to be placed in a facility. However, he feared nursing home admission represented giving up and failing to fulfill my duty to my mother. Tears ran down the face of this otherwise well composed, immaculately dressed, articulate man in his late forties. He had assumed the responsibility for his mother's care while still a child and had carried this self‐imposed moral burden his entire life. This meeting was his first opportunity to voice explicitly to the medical team his immense love and concern for his dear mother.

I gently probed to clarify Bertha's values and goals. On a brief, prior nursing home stay, Bertha had found the experience to be scary and unfamiliar. However, as her functional abilities continued to decline, her feelings had changed. She now felt lonely and anxious at home when her caretakers were absent. She actually wanted to go to a nursing home, where there would always be company and support available! She had not told Douglas this because she knew he cared for her deeply and she didn't want to hurt his feelings; he seemed committed to caring for her the same way he had for so many years.

In short, Douglas knew it would be easier for him if his mother were in a nursing facility, but assumed she wanted to stay in her home. Oddly, Bertha was only remaining at home because she believed that was what her son wished. A few days later, Bertha was transferred to a nursing home near several relatives who would visit her regularly. Douglas was again selfless in not seeking to move her closer to him. He didn't want to uproot her more than was unavoidable.

Day‐to‐day practice reveals many examples of love and dedication, but I have never seen such blinding and unquestioning commitment as exemplified by this mother‐son duo. From them, I learned the importance of attentive and active listening. Our polite patients may only subtly hint at matters of the deepest import. If we cannot truly hear their unspoken emotions, we risk harming them and misinterpreting their words and actions. Some healthcare providers had seen Douglas as aggressive and demanding with his threat of a lawsuit, whereas Bertha had been described as unrealistic or in denial. These views distorted a much more complex reality. Time and attention to careful communication between the healthcare providers, the patient, and her son bore fruit in this case. The procedure that was really needed was the family meeting and not the tracheostomy! An undesired and invasive procedure was avoided, goals of care were clarified, quality of life maximized, a safe discharge arranged, and a new mutual understanding achieved. I was humbled, and reminded of the importance of team‐based care and the need to approach each patient and family member in a receptive, nonjudgmental, and open manner. Douglas and Bertha Johnson were linked with a profound and abiding bond that would only be severed at death.

Bertha Johnson is back with pneumonia again. The ED doctor on the telephone sounded both matter‐of‐fact and mildly bored when I answered her page about another admission to the hospitalist service. I hadn't met Mrs. Johnson previously, but came to know her and Douglas, her only son, well over the next few days.

The Johnsons were facing a difficult choice. Bertha was now bedbound and quadriplegic following a 40‐year battle with multiple sclerosis and gradually mounting disability. She was cognitively intact and had a solid grasp of medical realities, but was hard of hearing and quite dysarthric. Forming even short phrases and sentences took great effort. However tenuous, this ability to speak allowed her to communicate with those she loved. She had been admitted thrice in the past year with aspiration pneumonia, as she was unable to clear her secretions reliably. Repeated bronchoscopies demonstrated an inability to protect her airway. Douglas, who was also her health care proxy, favored proceeding with the tracheostomy suggested by the pulmonary team. On a prior admission, he had been distressed when his mother refused this intervention. Now she was back with the identical problem and had been given the same recommendation by her doctors. It was particularly difficult for me to discuss these sensitive issues when I had not previously met either Bertha or her son. I spoke to her primary care physician over the phone and he agreed with the need for tracheostomy. The pulmonary team had been involved in discussions about tracheostomy in all of her hospital admissions, providing continuity of care in the process. Ultimately, it was my responsibility to help Bertha and Douglas come to a decision.

After multiple discussions between the Johnsons and me, a consensus emerged to proceed with the tracheostomy. We recognized that the procedure would increase her care needs and arranged for a stay in a skilled nursing facility to provide access to round‐the‐clock suctioning. The evening prior to the tracheostomy, the floor nurse and I reviewed the procedure to ensure that Bertha was fully prepared. What followed resulted in a drastic change of plan. Bertha emphasized that she did not want to lose her only means of communication, even if the surgery would prolong her life. She admitted that she reluctantly agreed to the procedure only to please her son and doctors, because they believed it to be in her best interest. Her fear of the prospect of death from drowning in her own secretions was much less than her fear of silence and isolation that would result from her loss of speech. I shared her misgivings with Douglas, and she admitted to him that she had only agreed to the procedure for his sake. We cancelled the surgery.

Douglas later revealed that he also had been ambivalent about the procedure for sometime, as it would necessitate a nursing home stay and the loss of the caretakers who had cared for his mother so wonderfully for many years. Bertha lived alone in her own home with the help of visiting nurses and patient care assistants Douglas paid for out‐of‐pocket. Douglas lived and worked in a city over a hundred miles away, but managed to visit several times a month to facilitate his mother's care. He had supported the procedure only because the doctors had said it was the only way to avoid future pneumonias. The idea of a tracheostomy was definitively abandoned once and for all.

The Johnsons wanted Bertha to return to her home, but hospital case managers felt this would be an unsafe plan of care as she was alone for several hours a day. She was largely immobile and unable to escape if there were a fire or other emergency. Also, the caretakers were not trained to use the suction equipment, and the visiting nurses would only be available intermittently. The home care staff felt they could no longer meet her needs and declined to resume her care. Douglas became very frustrated by the delays and protracted negotiations, enough so that he threatened to sue the institution for taking over my mother's life. The threat of litigation is usually a cry for help that reflects either miscommunication or the suffering of a conflicted family.

As their hospitalist, I hoped to advocate for both the patient and Douglas while coordinating the overall care plan. I had always received consistent responses from the Johnsons, but other staff members noted that Douglas had expressed shifting views on the best site of care for his mother. At Bertha's request, I convened a meeting with her, Douglas, the social worker, case manager, visiting and staff nurses, the palliative care nurse, and floor manager. Prior to this, I met with all involved health care providers to ensure we understood each other's abilities and limitations regarding Bertha's care. As I entered the room for the family meeting, I knew it was ultimately the patient's choice whether she wanted to return home or notas long as she understood the risks involved.

During the meeting, all the team members explained the dilemmas they faced in planning for a safe disposition. Douglas's response illuminated his devotion and love for his ailing mother. He had known all along that it would be less expensive and burdensome for him for Bertha to be placed in a facility. However, he feared nursing home admission represented giving up and failing to fulfill my duty to my mother. Tears ran down the face of this otherwise well composed, immaculately dressed, articulate man in his late forties. He had assumed the responsibility for his mother's care while still a child and had carried this self‐imposed moral burden his entire life. This meeting was his first opportunity to voice explicitly to the medical team his immense love and concern for his dear mother.

I gently probed to clarify Bertha's values and goals. On a brief, prior nursing home stay, Bertha had found the experience to be scary and unfamiliar. However, as her functional abilities continued to decline, her feelings had changed. She now felt lonely and anxious at home when her caretakers were absent. She actually wanted to go to a nursing home, where there would always be company and support available! She had not told Douglas this because she knew he cared for her deeply and she didn't want to hurt his feelings; he seemed committed to caring for her the same way he had for so many years.

In short, Douglas knew it would be easier for him if his mother were in a nursing facility, but assumed she wanted to stay in her home. Oddly, Bertha was only remaining at home because she believed that was what her son wished. A few days later, Bertha was transferred to a nursing home near several relatives who would visit her regularly. Douglas was again selfless in not seeking to move her closer to him. He didn't want to uproot her more than was unavoidable.

Day‐to‐day practice reveals many examples of love and dedication, but I have never seen such blinding and unquestioning commitment as exemplified by this mother‐son duo. From them, I learned the importance of attentive and active listening. Our polite patients may only subtly hint at matters of the deepest import. If we cannot truly hear their unspoken emotions, we risk harming them and misinterpreting their words and actions. Some healthcare providers had seen Douglas as aggressive and demanding with his threat of a lawsuit, whereas Bertha had been described as unrealistic or in denial. These views distorted a much more complex reality. Time and attention to careful communication between the healthcare providers, the patient, and her son bore fruit in this case. The procedure that was really needed was the family meeting and not the tracheostomy! An undesired and invasive procedure was avoided, goals of care were clarified, quality of life maximized, a safe discharge arranged, and a new mutual understanding achieved. I was humbled, and reminded of the importance of team‐based care and the need to approach each patient and family member in a receptive, nonjudgmental, and open manner. Douglas and Bertha Johnson were linked with a profound and abiding bond that would only be severed at death.

Bertha Johnson is back with pneumonia again. The ED doctor on the telephone sounded both matter‐of‐fact and mildly bored when I answered her page about another admission to the hospitalist service. I hadn't met Mrs. Johnson previously, but came to know her and Douglas, her only son, well over the next few days.

The Johnsons were facing a difficult choice. Bertha was now bedbound and quadriplegic following a 40‐year battle with multiple sclerosis and gradually mounting disability. She was cognitively intact and had a solid grasp of medical realities, but was hard of hearing and quite dysarthric. Forming even short phrases and sentences took great effort. However tenuous, this ability to speak allowed her to communicate with those she loved. She had been admitted thrice in the past year with aspiration pneumonia, as she was unable to clear her secretions reliably. Repeated bronchoscopies demonstrated an inability to protect her airway. Douglas, who was also her health care proxy, favored proceeding with the tracheostomy suggested by the pulmonary team. On a prior admission, he had been distressed when his mother refused this intervention. Now she was back with the identical problem and had been given the same recommendation by her doctors. It was particularly difficult for me to discuss these sensitive issues when I had not previously met either Bertha or her son. I spoke to her primary care physician over the phone and he agreed with the need for tracheostomy. The pulmonary team had been involved in discussions about tracheostomy in all of her hospital admissions, providing continuity of care in the process. Ultimately, it was my responsibility to help Bertha and Douglas come to a decision.

After multiple discussions between the Johnsons and me, a consensus emerged to proceed with the tracheostomy. We recognized that the procedure would increase her care needs and arranged for a stay in a skilled nursing facility to provide access to round‐the‐clock suctioning. The evening prior to the tracheostomy, the floor nurse and I reviewed the procedure to ensure that Bertha was fully prepared. What followed resulted in a drastic change of plan. Bertha emphasized that she did not want to lose her only means of communication, even if the surgery would prolong her life. She admitted that she reluctantly agreed to the procedure only to please her son and doctors, because they believed it to be in her best interest. Her fear of the prospect of death from drowning in her own secretions was much less than her fear of silence and isolation that would result from her loss of speech. I shared her misgivings with Douglas, and she admitted to him that she had only agreed to the procedure for his sake. We cancelled the surgery.

Douglas later revealed that he also had been ambivalent about the procedure for sometime, as it would necessitate a nursing home stay and the loss of the caretakers who had cared for his mother so wonderfully for many years. Bertha lived alone in her own home with the help of visiting nurses and patient care assistants Douglas paid for out‐of‐pocket. Douglas lived and worked in a city over a hundred miles away, but managed to visit several times a month to facilitate his mother's care. He had supported the procedure only because the doctors had said it was the only way to avoid future pneumonias. The idea of a tracheostomy was definitively abandoned once and for all.

The Johnsons wanted Bertha to return to her home, but hospital case managers felt this would be an unsafe plan of care as she was alone for several hours a day. She was largely immobile and unable to escape if there were a fire or other emergency. Also, the caretakers were not trained to use the suction equipment, and the visiting nurses would only be available intermittently. The home care staff felt they could no longer meet her needs and declined to resume her care. Douglas became very frustrated by the delays and protracted negotiations, enough so that he threatened to sue the institution for taking over my mother's life. The threat of litigation is usually a cry for help that reflects either miscommunication or the suffering of a conflicted family.

As their hospitalist, I hoped to advocate for both the patient and Douglas while coordinating the overall care plan. I had always received consistent responses from the Johnsons, but other staff members noted that Douglas had expressed shifting views on the best site of care for his mother. At Bertha's request, I convened a meeting with her, Douglas, the social worker, case manager, visiting and staff nurses, the palliative care nurse, and floor manager. Prior to this, I met with all involved health care providers to ensure we understood each other's abilities and limitations regarding Bertha's care. As I entered the room for the family meeting, I knew it was ultimately the patient's choice whether she wanted to return home or notas long as she understood the risks involved.

During the meeting, all the team members explained the dilemmas they faced in planning for a safe disposition. Douglas's response illuminated his devotion and love for his ailing mother. He had known all along that it would be less expensive and burdensome for him for Bertha to be placed in a facility. However, he feared nursing home admission represented giving up and failing to fulfill my duty to my mother. Tears ran down the face of this otherwise well composed, immaculately dressed, articulate man in his late forties. He had assumed the responsibility for his mother's care while still a child and had carried this self‐imposed moral burden his entire life. This meeting was his first opportunity to voice explicitly to the medical team his immense love and concern for his dear mother.

I gently probed to clarify Bertha's values and goals. On a brief, prior nursing home stay, Bertha had found the experience to be scary and unfamiliar. However, as her functional abilities continued to decline, her feelings had changed. She now felt lonely and anxious at home when her caretakers were absent. She actually wanted to go to a nursing home, where there would always be company and support available! She had not told Douglas this because she knew he cared for her deeply and she didn't want to hurt his feelings; he seemed committed to caring for her the same way he had for so many years.

In short, Douglas knew it would be easier for him if his mother were in a nursing facility, but assumed she wanted to stay in her home. Oddly, Bertha was only remaining at home because she believed that was what her son wished. A few days later, Bertha was transferred to a nursing home near several relatives who would visit her regularly. Douglas was again selfless in not seeking to move her closer to him. He didn't want to uproot her more than was unavoidable.

Day‐to‐day practice reveals many examples of love and dedication, but I have never seen such blinding and unquestioning commitment as exemplified by this mother‐son duo. From them, I learned the importance of attentive and active listening. Our polite patients may only subtly hint at matters of the deepest import. If we cannot truly hear their unspoken emotions, we risk harming them and misinterpreting their words and actions. Some healthcare providers had seen Douglas as aggressive and demanding with his threat of a lawsuit, whereas Bertha had been described as unrealistic or in denial. These views distorted a much more complex reality. Time and attention to careful communication between the healthcare providers, the patient, and her son bore fruit in this case. The procedure that was really needed was the family meeting and not the tracheostomy! An undesired and invasive procedure was avoided, goals of care were clarified, quality of life maximized, a safe discharge arranged, and a new mutual understanding achieved. I was humbled, and reminded of the importance of team‐based care and the need to approach each patient and family member in a receptive, nonjudgmental, and open manner. Douglas and Bertha Johnson were linked with a profound and abiding bond that would only be severed at death.

A 19‐year‐old Japanese man was admitted to a hospital near Kyoto, Japan, because of fever and rash. Two weeks prior to admission, he developed mild headache and low‐grade fever; a rapid test for influenza was negative. His symptoms transiently improved with acetaminophen, but 8 days prior to admission, he developed fever to 38.5C and a pruritic maculopapular rash over his back that spread to his limbs. Six days prior to admission, a chest radiograph was clear; clarithromycin was prescribed for presumed upper respiratory infection. He visited the emergency department the day before admission because of continued fever of greater than 39C, fatigue, and headache. Because there was no jolt accentuation of the headache (ie, worsening with rapid horizontal rotation), or neck pain with extreme neck flexion, he was discharged on acetaminophen. He returned the next day with worsening fatigue and was admitted. He denied chills, rigor, weight loss, photosensitivity, sore throat, neck pain, cough, dyspnea, chest pain, nausea, vomiting, diarrhea, abdominal pain, back pain, and arthralgia.

Fever and diffuse rash are often due to infection, although drugs, autoimmune processes, and cancer must be considered. The presence of headache does not focus the differential diagnosis substantially, because many of the candidate diagnoses can be accompanied by meningitis or encephalitis, or even more frequently, nonspecific headaches. In one small study, jolt‐induced aggravation of headache was shown to be a sensitive indicator of cerebrospinal fluid pleocytosis. The absence of neck stiffness and the 2‐week duration makes bacterial meningitis unlikely, but a more indolent form of aseptic meningitis may need to be evaluated with a lumbar puncture.

The 2‐week illness without rapid deterioration makes some serious causes of fever and rash, such as toxic shock syndrome, disseminated meningococcal infection, or toxic epidermal necrolysis unlikely. A viral exanthema is possible, although the 2‐week duration is longer than usual. Given his youth, however, his immunization history should be queried, and acute infection with human immunodeficiency virus (HIV) should be considered. A more indolent infection, such as subacute bacterial endocarditis, disseminated gonococcal infection, or syphilis is plausible. Among autoimmune etiologies, systemic lupus erythematosus (SLE) and Behcet's disease (which is prevalent in Japan) can involve the central nervous system and cause fever. A careful inquiry directed at prescribed, complementary, and illicit drugs is required.

The patient's past medical history was notable only for mumps at the age of 10. His medications included acetaminophen, clarithromycin, and an herbal medicine, which he had been taking for the prior several days. He reported no tobacco or illicit drug use and rarely drank alcohol. He had never been sexually active. He worked in a factory and reported occasional contact with silver. He lived with his parents; there was no family history of tuberculosis or connective tissue diseases. His father was from Kyushu (the southernmost major island in Japan) and had chronic hepatitis C. The patient denied recent animal exposure or recent travel. His childhood vaccinations were said to be up to date.

Mumps at age 10 might signal general lack of immunization, in which case childhood viral exanthema‐like measles (characterized by fever, headache, and diffuse rash) would warrant consideration. The listed medications had been started after the onset of illness and therefore are unlikely to be causal. Silver causes at least 2 skin conditionscontact dermatitis and argyriabut not the systemic illness seen here. Human T lymphotropic virus‐1 (HTLV‐1) is endemic in southern Japan, but only a minority of infected humans are afflicted with associated adult T cell leukemia/lymphoma or myelopathy. Leukemia and lymphoma are the most likely cancers to cause fever, rash, and central nervous system involvement (with T cell disorders demonstrating a particular tropism for the skin). Overall, however, the differential has not changed substantially.

On physical examination, the patient was mildly overweight and appeared acutely ill. His blood pressure was 136/78 mm Hg, pulse rate was 76 and regular, temperature was 39.2C and respiratory rate was 20 with an oxygen saturation of 98% on room air. A diffuse but nonconfluent erythematous maculopapular rash was present over his chest wall, back, medial aspects of both thighs, and around the knees. There was no jolt‐induced headache. His eyes, nose, oral cavity, and throat were all clear. The neck was supple. There were palpable lymph nodes, each about 1 cm in size, which were firm and moderately tender, in his left neck and left axilla. Lungs and heart were normal. The abdomen was soft, nontender, with normal bowel sounds and no hepatosplenomegaly. His genitalia were normal. Rectal examination revealed no masses or tenderness and a scant amount of brown stool that was negative for occult blood. Neurologic examination was unremarkable.

The multifocal lymphadenopathy does not help distinguish among the categories of disease under consideration. The diffuse maculopapular rash is similarly nonspecific, occurring more frequently with infection and drug reaction than malignancy and autoimmunity. Acute HIV, Epstein‐Barr virus (EBV), syphilis, SLE, drug exposure, or a hematologic malignancy would all be suitable explanations for fever, headache, diffuse rash, and disseminated lymphadenopathy in a previously healthy young man.

Laboratory data obtained on admission was notable for a white blood cell (WBC) count of 2100/L with 72% neutrophils, 19% lymphocytes, and 9% monocytes. Hemoglobin was 13.5 mg/dL with a mean corpuscular volume of 85 fL. Platelet count was 136,000/L. Erythrocyte sedimentation rate was 26 mm/hour. Serum chemistries revealed a sodium level of 135 mEq/L, potassium level of 3.6 mEq/L, chloride level of 100 mEq/L, blood urea nitrogen of 9.8 mg/dL, creatinine level of 1.0 mg/dL, glucose level of 101 mg/dL, calcium level of 8.8 mg/dL, albumin of 4.6 mg/dL, total protein of 8.4 mg/dL, aspartate aminotransferase of 42 IU/L (normal < 35 IU/L), alanine aminotransferase of 27 IU/L, total bilirubin of 0.5 mg/dL, and lactate dehydrogenase (LDH) level of 463 IU/L (normal < 260 IU/L). Chest radiography and electrocardiogram were normal.

A mild elevation in LDH is nonspecific, but without hemolysis or infarction of the kidney, lung, or muscle, it suggests a lymphoproliferative process. Leukopenia with thrombocytopenia can be seen in a number of disorders, most commonly infections including viruses (e.g., EBV, HIV, dengue), malaria, Rocky Mountain spotted fever, or ehrlichiosis/anaplasmosis. Confirmation of his lack of travel could help prioritize those considerations. An invasive bone marrow disorder cannot be excluded, although the near‐normal hemoglobin argues against it. Autoimmune cytopenias are seen in SLE. Given his age, lymphadenopathy, LDH elevation, and absence of infectious exposures, lymphoma rises to the top of the list.

Noninvasive measures should include examination of the peripheral smear, HIV testing (including HIV RNA for acute infection), EBV serologies, and tests for syphilis and SLE. Lumbar puncture (for evaluation of aseptic meningitis) and lymph node biopsy would be informative. Skin biopsy may be helpful to evaluate for aggressive T cell lymphoproliferative disorder, but this can await the results of initial testing.

The patient was given intravenous fluids and acetaminophen as needed. Blood cultures, urine culture, cytomegalovirus and EBV serologies, hepatitis B surface antigen, hepatitis C virus antibody, HIV antibody, antinuclear antibody, complement and ferritin levels, and quantiferon‐TB were ordered. The urine was normal and a urinary antigen test for Legionella was negative. Contrast‐enhanced computed tomography scan of the chest and abdomen was normal except for mild splenomegaly and an enlarged left axillary lymph node.

The ferritin may have been ordered to help evaluate for Still's disease, which is characterized by sustained fever, lymphadenopathy, and transient rash; however, the characteristic leukocytosis and arthralgias are absent. The computed tomography findings are most notable for the absence of generalized lymphadenopathy or significant hepatosplenomegaly that is seen in lymphoma, leukemia, and lymphotropic processes such as acute EBV infection. The localization of disease to the skin (where the predominant lymphocytes are of T cell origin) with relatively modest lymphadenopathy suggests a T cell lymphoma, perhaps of an indolent variety. Vertical transmission of HTLV‐1 decades ago would make adult T cell leukemia or lymphoma a major consideration.

On the third hospital day, WBC count was 1800/L with 67% neutrophils, 22% lymphocytes, and 1% atypical lymphocytes; LDH rose to 623 IU/L. He had continued fatigue and high fever while the rash gradually faded with oral antihistamines and steroid ointment. On hospital day 4, bone marrow biopsy and skin biopsy of his left thigh were performed.

The further decline in WBC and rise in LDH are modest and therefore do not significantly modify the differential diagnosis. Likewise, 1% atypical lymphocytosis is too low to pinpoint an etiology. Because unremitting fevers start to extend into their third week without a clear source of infection, the probability of malignancy and autoimmunity rise. Improvement with oral antihistamines and topical steroids frequently suggests an underlying allergic process, but the remainder of the clinical picture is not in keeping with atopy or allergy. Cutaneous lymphomas (eg, mycosis fungoides) can have waxing and waning skin manifestations, and can be temporarily or definitively treated by topical steroids. The persistence of his fatigue is of concern given the absence of anemia, cardiopulmonary involvement, or motor weakness.

Bone marrow biopsy showed normocellular marrow with no abnormal cells and some activated macrophages with hemophagocytic activity. Skin biopsy failed to show specific pathology.

His left cervical lymph nodes gradually enlarged. Ultrasound of the neck showed multiple enlarged lymph nodes (left side dominant) with dimension of 17 mm 9 mm 31 mm. Blood and urine cultures returned negative, as did HIV antibody. cytomegalovirus and EBV serologies were consistent with previous infection and the ferritin level was 578 ng/mL (normal, 39‐340 ng/mL). Toxoplasma serology and HTLV‐1 antibody were ordered.

The absence of malignant cells on bone marrow biopsy does not exclude lymphoma, but makes a myelophthisic cause of the cytopenias less likely. The macrophage hemophagocytosis reflects immune activation, which in turn is usually caused by the same viral infections, autoimmune conditions, and lymphoproliferative disorders which constitute the current differential diagnosis.

Bone marrow and skin biopsies are both subject to sampling error, and detection of cutaneous T cell lymphoma is notoriously difficult. However, taken together, the absence of cancer on 2 specimens reduces that possibility.

Sustained unilateral cervical lymphadenopathy with fever in a young Japanese man without any histologic evidence of lymphoma points to Kikuchi's disease, ie, lymphadenitis of unknown etiology associated with varying degrees of systemic manifestations. Fever is a frequent feature, we believe, but diffuse sustained rash, cytopenias, and headache are less common or are seen in severe forms of the disease. The diagnosis of Kikuchi's requires the diligent exclusion of SLE and lymphoma. Examination of the peripheral smear and a lymph node biopsy are required.

Of note, there is also a localized form of Castleman's disease, a nonmalignant lymphoproliferative disorder, that similarly is characterized by focal lymphadenopathy. In distinction to Kikuchi's, however, localized Castleman's is largely asymptomatic and responds marvelously to excision.

On hospital day 9, an excisional biopsy of his left anterior cervical lymph nodes was performed, which revealed paracortical foci with necrosis and a histiocytic cellular infiltrate consistent with subacute necrotizing lymphadenitis (Kikuchi‐Fujimoto disease). Antinuclear antibody, Toxoplasma, and HTLV‐1 antibodies returned negative.

There is no treatment for Kikuchi's. It is usually self‐limited, but steroids are sometimes given for symptomatic control.

His condition began to improve after hospital day 9 without specific treatment, including his WBC count and LDH level. He was discharged home on hospital day 15. In the outpatient clinic 1 and 3 months later, he was well and active without recurrences of any symptoms or laboratory abnormalities. His WBC count was 6600/L and LDH was 268 IU/L.

Commentary

Kikuchi‐Fujimoto disease (KFD), also called Kikuchi's disease, is a benign histiocytic necrotizing lymphadenitis described by both Kikuchi and Fujimoto in 1972.1, 2 It is rare in the United States, but seems more common in Asia, especially Japan, where at least 143 cases have been reported since 1972. The etiology has not been determined, but a viral causeincluding EBV, and human herpesvirus 6 and 8has been suggested.3 An autoimmune etiology is also implicated because of infrequent association with SLE. In general, young women are most likely to be affected. In a review of 244 cases by Kucukardali and colleagues, 77% of patients were female and the mean age was 25; 70% were younger than 30 years of age.4

The common presentation is low‐grade fever with unilateral cervical lymphadenopathy.4 Although generalized lymphadenopathy can occur, it is rare. Other common clinical manifestations include malaise, joint pain, rash, arthritis, and hepatosplenomegaly. No specific laboratory tests for diagnosis are available, but leukopenia (seen in 43% of patients), increased erythrocyte sedimentation rate (40%), and anemia (23%) may be observed.4 In this case, atypical lymphocytes were seen, and are reported in one‐third of patients.5 KFD is generally diagnosed by lymph node biopsy, which typically shows irregular paracortical areas of coagulation necrosis that can distort the nodal architecture, while different types of histiocytes are observed at the margin of necrotic areas.

Other diseases in the differential diagnosisseveral of which were considered by the discussantinclude lymphoma, tuberculosis, SLE, and even metastatic adenocarcinoma. KFD is self‐limited; symptoms typically resolve within 1 to 4 months. Patients with severe manifestations have been treated with anti‐inflammatory drugs and glucocorticosteroids. A recurrence rate of 3% to 4% has been reported.6

The clinicians taking care of this patient initially focused on ruling out those infections occasionally resulting in prolonged fever in a previously healthy young man, such as viruses from the herpes family, HIV, viral hepatitis, tuberculosis, syphilis, infective endocarditis, and intra‐abdominal abscess. Physical examination, specifically lymphadenopathy and mild splenomegaly, made Herpesviridae infections, tuberculosis, syphilis, and lymphoma difficult to exclude. Once the initial evaluation ruled out common infections, attention focused on malignancy and histiocytic necrotizing lymphadenitis, given his ethnicity and geographic location.

The discussant was similarly concerned about infection, malignancy, and noninfectious inflammatory diseases, such as SLE, as possible causes. As evidence of these treatable diseases failed to accumulate, the discussant, an American physician with teaching and clinical experience in Japan, considered endemic diseases such as Behcet's, HTLV‐1, and KFD because they fit the unfolding pattern. Given our global society, clinicians will increasingly benefit from becoming familiar with the less common diseases that afflict the various populations around the world.

A 19‐year‐old Japanese man was admitted to a hospital near Kyoto, Japan, because of fever and rash. Two weeks prior to admission, he developed mild headache and low‐grade fever; a rapid test for influenza was negative. His symptoms transiently improved with acetaminophen, but 8 days prior to admission, he developed fever to 38.5C and a pruritic maculopapular rash over his back that spread to his limbs. Six days prior to admission, a chest radiograph was clear; clarithromycin was prescribed for presumed upper respiratory infection. He visited the emergency department the day before admission because of continued fever of greater than 39C, fatigue, and headache. Because there was no jolt accentuation of the headache (ie, worsening with rapid horizontal rotation), or neck pain with extreme neck flexion, he was discharged on acetaminophen. He returned the next day with worsening fatigue and was admitted. He denied chills, rigor, weight loss, photosensitivity, sore throat, neck pain, cough, dyspnea, chest pain, nausea, vomiting, diarrhea, abdominal pain, back pain, and arthralgia.

Fever and diffuse rash are often due to infection, although drugs, autoimmune processes, and cancer must be considered. The presence of headache does not focus the differential diagnosis substantially, because many of the candidate diagnoses can be accompanied by meningitis or encephalitis, or even more frequently, nonspecific headaches. In one small study, jolt‐induced aggravation of headache was shown to be a sensitive indicator of cerebrospinal fluid pleocytosis. The absence of neck stiffness and the 2‐week duration makes bacterial meningitis unlikely, but a more indolent form of aseptic meningitis may need to be evaluated with a lumbar puncture.

The 2‐week illness without rapid deterioration makes some serious causes of fever and rash, such as toxic shock syndrome, disseminated meningococcal infection, or toxic epidermal necrolysis unlikely. A viral exanthema is possible, although the 2‐week duration is longer than usual. Given his youth, however, his immunization history should be queried, and acute infection with human immunodeficiency virus (HIV) should be considered. A more indolent infection, such as subacute bacterial endocarditis, disseminated gonococcal infection, or syphilis is plausible. Among autoimmune etiologies, systemic lupus erythematosus (SLE) and Behcet's disease (which is prevalent in Japan) can involve the central nervous system and cause fever. A careful inquiry directed at prescribed, complementary, and illicit drugs is required.

The patient's past medical history was notable only for mumps at the age of 10. His medications included acetaminophen, clarithromycin, and an herbal medicine, which he had been taking for the prior several days. He reported no tobacco or illicit drug use and rarely drank alcohol. He had never been sexually active. He worked in a factory and reported occasional contact with silver. He lived with his parents; there was no family history of tuberculosis or connective tissue diseases. His father was from Kyushu (the southernmost major island in Japan) and had chronic hepatitis C. The patient denied recent animal exposure or recent travel. His childhood vaccinations were said to be up to date.

Mumps at age 10 might signal general lack of immunization, in which case childhood viral exanthema‐like measles (characterized by fever, headache, and diffuse rash) would warrant consideration. The listed medications had been started after the onset of illness and therefore are unlikely to be causal. Silver causes at least 2 skin conditionscontact dermatitis and argyriabut not the systemic illness seen here. Human T lymphotropic virus‐1 (HTLV‐1) is endemic in southern Japan, but only a minority of infected humans are afflicted with associated adult T cell leukemia/lymphoma or myelopathy. Leukemia and lymphoma are the most likely cancers to cause fever, rash, and central nervous system involvement (with T cell disorders demonstrating a particular tropism for the skin). Overall, however, the differential has not changed substantially.

On physical examination, the patient was mildly overweight and appeared acutely ill. His blood pressure was 136/78 mm Hg, pulse rate was 76 and regular, temperature was 39.2C and respiratory rate was 20 with an oxygen saturation of 98% on room air. A diffuse but nonconfluent erythematous maculopapular rash was present over his chest wall, back, medial aspects of both thighs, and around the knees. There was no jolt‐induced headache. His eyes, nose, oral cavity, and throat were all clear. The neck was supple. There were palpable lymph nodes, each about 1 cm in size, which were firm and moderately tender, in his left neck and left axilla. Lungs and heart were normal. The abdomen was soft, nontender, with normal bowel sounds and no hepatosplenomegaly. His genitalia were normal. Rectal examination revealed no masses or tenderness and a scant amount of brown stool that was negative for occult blood. Neurologic examination was unremarkable.

The multifocal lymphadenopathy does not help distinguish among the categories of disease under consideration. The diffuse maculopapular rash is similarly nonspecific, occurring more frequently with infection and drug reaction than malignancy and autoimmunity. Acute HIV, Epstein‐Barr virus (EBV), syphilis, SLE, drug exposure, or a hematologic malignancy would all be suitable explanations for fever, headache, diffuse rash, and disseminated lymphadenopathy in a previously healthy young man.

Laboratory data obtained on admission was notable for a white blood cell (WBC) count of 2100/L with 72% neutrophils, 19% lymphocytes, and 9% monocytes. Hemoglobin was 13.5 mg/dL with a mean corpuscular volume of 85 fL. Platelet count was 136,000/L. Erythrocyte sedimentation rate was 26 mm/hour. Serum chemistries revealed a sodium level of 135 mEq/L, potassium level of 3.6 mEq/L, chloride level of 100 mEq/L, blood urea nitrogen of 9.8 mg/dL, creatinine level of 1.0 mg/dL, glucose level of 101 mg/dL, calcium level of 8.8 mg/dL, albumin of 4.6 mg/dL, total protein of 8.4 mg/dL, aspartate aminotransferase of 42 IU/L (normal < 35 IU/L), alanine aminotransferase of 27 IU/L, total bilirubin of 0.5 mg/dL, and lactate dehydrogenase (LDH) level of 463 IU/L (normal < 260 IU/L). Chest radiography and electrocardiogram were normal.

A mild elevation in LDH is nonspecific, but without hemolysis or infarction of the kidney, lung, or muscle, it suggests a lymphoproliferative process. Leukopenia with thrombocytopenia can be seen in a number of disorders, most commonly infections including viruses (e.g., EBV, HIV, dengue), malaria, Rocky Mountain spotted fever, or ehrlichiosis/anaplasmosis. Confirmation of his lack of travel could help prioritize those considerations. An invasive bone marrow disorder cannot be excluded, although the near‐normal hemoglobin argues against it. Autoimmune cytopenias are seen in SLE. Given his age, lymphadenopathy, LDH elevation, and absence of infectious exposures, lymphoma rises to the top of the list.

Noninvasive measures should include examination of the peripheral smear, HIV testing (including HIV RNA for acute infection), EBV serologies, and tests for syphilis and SLE. Lumbar puncture (for evaluation of aseptic meningitis) and lymph node biopsy would be informative. Skin biopsy may be helpful to evaluate for aggressive T cell lymphoproliferative disorder, but this can await the results of initial testing.

The patient was given intravenous fluids and acetaminophen as needed. Blood cultures, urine culture, cytomegalovirus and EBV serologies, hepatitis B surface antigen, hepatitis C virus antibody, HIV antibody, antinuclear antibody, complement and ferritin levels, and quantiferon‐TB were ordered. The urine was normal and a urinary antigen test for Legionella was negative. Contrast‐enhanced computed tomography scan of the chest and abdomen was normal except for mild splenomegaly and an enlarged left axillary lymph node.

The ferritin may have been ordered to help evaluate for Still's disease, which is characterized by sustained fever, lymphadenopathy, and transient rash; however, the characteristic leukocytosis and arthralgias are absent. The computed tomography findings are most notable for the absence of generalized lymphadenopathy or significant hepatosplenomegaly that is seen in lymphoma, leukemia, and lymphotropic processes such as acute EBV infection. The localization of disease to the skin (where the predominant lymphocytes are of T cell origin) with relatively modest lymphadenopathy suggests a T cell lymphoma, perhaps of an indolent variety. Vertical transmission of HTLV‐1 decades ago would make adult T cell leukemia or lymphoma a major consideration.

On the third hospital day, WBC count was 1800/L with 67% neutrophils, 22% lymphocytes, and 1% atypical lymphocytes; LDH rose to 623 IU/L. He had continued fatigue and high fever while the rash gradually faded with oral antihistamines and steroid ointment. On hospital day 4, bone marrow biopsy and skin biopsy of his left thigh were performed.

The further decline in WBC and rise in LDH are modest and therefore do not significantly modify the differential diagnosis. Likewise, 1% atypical lymphocytosis is too low to pinpoint an etiology. Because unremitting fevers start to extend into their third week without a clear source of infection, the probability of malignancy and autoimmunity rise. Improvement with oral antihistamines and topical steroids frequently suggests an underlying allergic process, but the remainder of the clinical picture is not in keeping with atopy or allergy. Cutaneous lymphomas (eg, mycosis fungoides) can have waxing and waning skin manifestations, and can be temporarily or definitively treated by topical steroids. The persistence of his fatigue is of concern given the absence of anemia, cardiopulmonary involvement, or motor weakness.

Bone marrow biopsy showed normocellular marrow with no abnormal cells and some activated macrophages with hemophagocytic activity. Skin biopsy failed to show specific pathology.

His left cervical lymph nodes gradually enlarged. Ultrasound of the neck showed multiple enlarged lymph nodes (left side dominant) with dimension of 17 mm 9 mm 31 mm. Blood and urine cultures returned negative, as did HIV antibody. cytomegalovirus and EBV serologies were consistent with previous infection and the ferritin level was 578 ng/mL (normal, 39‐340 ng/mL). Toxoplasma serology and HTLV‐1 antibody were ordered.

The absence of malignant cells on bone marrow biopsy does not exclude lymphoma, but makes a myelophthisic cause of the cytopenias less likely. The macrophage hemophagocytosis reflects immune activation, which in turn is usually caused by the same viral infections, autoimmune conditions, and lymphoproliferative disorders which constitute the current differential diagnosis.

Bone marrow and skin biopsies are both subject to sampling error, and detection of cutaneous T cell lymphoma is notoriously difficult. However, taken together, the absence of cancer on 2 specimens reduces that possibility.

Sustained unilateral cervical lymphadenopathy with fever in a young Japanese man without any histologic evidence of lymphoma points to Kikuchi's disease, ie, lymphadenitis of unknown etiology associated with varying degrees of systemic manifestations. Fever is a frequent feature, we believe, but diffuse sustained rash, cytopenias, and headache are less common or are seen in severe forms of the disease. The diagnosis of Kikuchi's requires the diligent exclusion of SLE and lymphoma. Examination of the peripheral smear and a lymph node biopsy are required.

Of note, there is also a localized form of Castleman's disease, a nonmalignant lymphoproliferative disorder, that similarly is characterized by focal lymphadenopathy. In distinction to Kikuchi's, however, localized Castleman's is largely asymptomatic and responds marvelously to excision.

On hospital day 9, an excisional biopsy of his left anterior cervical lymph nodes was performed, which revealed paracortical foci with necrosis and a histiocytic cellular infiltrate consistent with subacute necrotizing lymphadenitis (Kikuchi‐Fujimoto disease). Antinuclear antibody, Toxoplasma, and HTLV‐1 antibodies returned negative.

There is no treatment for Kikuchi's. It is usually self‐limited, but steroids are sometimes given for symptomatic control.

His condition began to improve after hospital day 9 without specific treatment, including his WBC count and LDH level. He was discharged home on hospital day 15. In the outpatient clinic 1 and 3 months later, he was well and active without recurrences of any symptoms or laboratory abnormalities. His WBC count was 6600/L and LDH was 268 IU/L.

Commentary

Kikuchi‐Fujimoto disease (KFD), also called Kikuchi's disease, is a benign histiocytic necrotizing lymphadenitis described by both Kikuchi and Fujimoto in 1972.1, 2 It is rare in the United States, but seems more common in Asia, especially Japan, where at least 143 cases have been reported since 1972. The etiology has not been determined, but a viral causeincluding EBV, and human herpesvirus 6 and 8has been suggested.3 An autoimmune etiology is also implicated because of infrequent association with SLE. In general, young women are most likely to be affected. In a review of 244 cases by Kucukardali and colleagues, 77% of patients were female and the mean age was 25; 70% were younger than 30 years of age.4

The common presentation is low‐grade fever with unilateral cervical lymphadenopathy.4 Although generalized lymphadenopathy can occur, it is rare. Other common clinical manifestations include malaise, joint pain, rash, arthritis, and hepatosplenomegaly. No specific laboratory tests for diagnosis are available, but leukopenia (seen in 43% of patients), increased erythrocyte sedimentation rate (40%), and anemia (23%) may be observed.4 In this case, atypical lymphocytes were seen, and are reported in one‐third of patients.5 KFD is generally diagnosed by lymph node biopsy, which typically shows irregular paracortical areas of coagulation necrosis that can distort the nodal architecture, while different types of histiocytes are observed at the margin of necrotic areas.

Other diseases in the differential diagnosisseveral of which were considered by the discussantinclude lymphoma, tuberculosis, SLE, and even metastatic adenocarcinoma. KFD is self‐limited; symptoms typically resolve within 1 to 4 months. Patients with severe manifestations have been treated with anti‐inflammatory drugs and glucocorticosteroids. A recurrence rate of 3% to 4% has been reported.6

The clinicians taking care of this patient initially focused on ruling out those infections occasionally resulting in prolonged fever in a previously healthy young man, such as viruses from the herpes family, HIV, viral hepatitis, tuberculosis, syphilis, infective endocarditis, and intra‐abdominal abscess. Physical examination, specifically lymphadenopathy and mild splenomegaly, made Herpesviridae infections, tuberculosis, syphilis, and lymphoma difficult to exclude. Once the initial evaluation ruled out common infections, attention focused on malignancy and histiocytic necrotizing lymphadenitis, given his ethnicity and geographic location.

The discussant was similarly concerned about infection, malignancy, and noninfectious inflammatory diseases, such as SLE, as possible causes. As evidence of these treatable diseases failed to accumulate, the discussant, an American physician with teaching and clinical experience in Japan, considered endemic diseases such as Behcet's, HTLV‐1, and KFD because they fit the unfolding pattern. Given our global society, clinicians will increasingly benefit from becoming familiar with the less common diseases that afflict the various populations around the world.

A 19‐year‐old Japanese man was admitted to a hospital near Kyoto, Japan, because of fever and rash. Two weeks prior to admission, he developed mild headache and low‐grade fever; a rapid test for influenza was negative. His symptoms transiently improved with acetaminophen, but 8 days prior to admission, he developed fever to 38.5C and a pruritic maculopapular rash over his back that spread to his limbs. Six days prior to admission, a chest radiograph was clear; clarithromycin was prescribed for presumed upper respiratory infection. He visited the emergency department the day before admission because of continued fever of greater than 39C, fatigue, and headache. Because there was no jolt accentuation of the headache (ie, worsening with rapid horizontal rotation), or neck pain with extreme neck flexion, he was discharged on acetaminophen. He returned the next day with worsening fatigue and was admitted. He denied chills, rigor, weight loss, photosensitivity, sore throat, neck pain, cough, dyspnea, chest pain, nausea, vomiting, diarrhea, abdominal pain, back pain, and arthralgia.

Fever and diffuse rash are often due to infection, although drugs, autoimmune processes, and cancer must be considered. The presence of headache does not focus the differential diagnosis substantially, because many of the candidate diagnoses can be accompanied by meningitis or encephalitis, or even more frequently, nonspecific headaches. In one small study, jolt‐induced aggravation of headache was shown to be a sensitive indicator of cerebrospinal fluid pleocytosis. The absence of neck stiffness and the 2‐week duration makes bacterial meningitis unlikely, but a more indolent form of aseptic meningitis may need to be evaluated with a lumbar puncture.

The 2‐week illness without rapid deterioration makes some serious causes of fever and rash, such as toxic shock syndrome, disseminated meningococcal infection, or toxic epidermal necrolysis unlikely. A viral exanthema is possible, although the 2‐week duration is longer than usual. Given his youth, however, his immunization history should be queried, and acute infection with human immunodeficiency virus (HIV) should be considered. A more indolent infection, such as subacute bacterial endocarditis, disseminated gonococcal infection, or syphilis is plausible. Among autoimmune etiologies, systemic lupus erythematosus (SLE) and Behcet's disease (which is prevalent in Japan) can involve the central nervous system and cause fever. A careful inquiry directed at prescribed, complementary, and illicit drugs is required.

The patient's past medical history was notable only for mumps at the age of 10. His medications included acetaminophen, clarithromycin, and an herbal medicine, which he had been taking for the prior several days. He reported no tobacco or illicit drug use and rarely drank alcohol. He had never been sexually active. He worked in a factory and reported occasional contact with silver. He lived with his parents; there was no family history of tuberculosis or connective tissue diseases. His father was from Kyushu (the southernmost major island in Japan) and had chronic hepatitis C. The patient denied recent animal exposure or recent travel. His childhood vaccinations were said to be up to date.

Mumps at age 10 might signal general lack of immunization, in which case childhood viral exanthema‐like measles (characterized by fever, headache, and diffuse rash) would warrant consideration. The listed medications had been started after the onset of illness and therefore are unlikely to be causal. Silver causes at least 2 skin conditionscontact dermatitis and argyriabut not the systemic illness seen here. Human T lymphotropic virus‐1 (HTLV‐1) is endemic in southern Japan, but only a minority of infected humans are afflicted with associated adult T cell leukemia/lymphoma or myelopathy. Leukemia and lymphoma are the most likely cancers to cause fever, rash, and central nervous system involvement (with T cell disorders demonstrating a particular tropism for the skin). Overall, however, the differential has not changed substantially.

On physical examination, the patient was mildly overweight and appeared acutely ill. His blood pressure was 136/78 mm Hg, pulse rate was 76 and regular, temperature was 39.2C and respiratory rate was 20 with an oxygen saturation of 98% on room air. A diffuse but nonconfluent erythematous maculopapular rash was present over his chest wall, back, medial aspects of both thighs, and around the knees. There was no jolt‐induced headache. His eyes, nose, oral cavity, and throat were all clear. The neck was supple. There were palpable lymph nodes, each about 1 cm in size, which were firm and moderately tender, in his left neck and left axilla. Lungs and heart were normal. The abdomen was soft, nontender, with normal bowel sounds and no hepatosplenomegaly. His genitalia were normal. Rectal examination revealed no masses or tenderness and a scant amount of brown stool that was negative for occult blood. Neurologic examination was unremarkable.

The multifocal lymphadenopathy does not help distinguish among the categories of disease under consideration. The diffuse maculopapular rash is similarly nonspecific, occurring more frequently with infection and drug reaction than malignancy and autoimmunity. Acute HIV, Epstein‐Barr virus (EBV), syphilis, SLE, drug exposure, or a hematologic malignancy would all be suitable explanations for fever, headache, diffuse rash, and disseminated lymphadenopathy in a previously healthy young man.

Laboratory data obtained on admission was notable for a white blood cell (WBC) count of 2100/L with 72% neutrophils, 19% lymphocytes, and 9% monocytes. Hemoglobin was 13.5 mg/dL with a mean corpuscular volume of 85 fL. Platelet count was 136,000/L. Erythrocyte sedimentation rate was 26 mm/hour. Serum chemistries revealed a sodium level of 135 mEq/L, potassium level of 3.6 mEq/L, chloride level of 100 mEq/L, blood urea nitrogen of 9.8 mg/dL, creatinine level of 1.0 mg/dL, glucose level of 101 mg/dL, calcium level of 8.8 mg/dL, albumin of 4.6 mg/dL, total protein of 8.4 mg/dL, aspartate aminotransferase of 42 IU/L (normal < 35 IU/L), alanine aminotransferase of 27 IU/L, total bilirubin of 0.5 mg/dL, and lactate dehydrogenase (LDH) level of 463 IU/L (normal < 260 IU/L). Chest radiography and electrocardiogram were normal.

A mild elevation in LDH is nonspecific, but without hemolysis or infarction of the kidney, lung, or muscle, it suggests a lymphoproliferative process. Leukopenia with thrombocytopenia can be seen in a number of disorders, most commonly infections including viruses (e.g., EBV, HIV, dengue), malaria, Rocky Mountain spotted fever, or ehrlichiosis/anaplasmosis. Confirmation of his lack of travel could help prioritize those considerations. An invasive bone marrow disorder cannot be excluded, although the near‐normal hemoglobin argues against it. Autoimmune cytopenias are seen in SLE. Given his age, lymphadenopathy, LDH elevation, and absence of infectious exposures, lymphoma rises to the top of the list.

Noninvasive measures should include examination of the peripheral smear, HIV testing (including HIV RNA for acute infection), EBV serologies, and tests for syphilis and SLE. Lumbar puncture (for evaluation of aseptic meningitis) and lymph node biopsy would be informative. Skin biopsy may be helpful to evaluate for aggressive T cell lymphoproliferative disorder, but this can await the results of initial testing.

The patient was given intravenous fluids and acetaminophen as needed. Blood cultures, urine culture, cytomegalovirus and EBV serologies, hepatitis B surface antigen, hepatitis C virus antibody, HIV antibody, antinuclear antibody, complement and ferritin levels, and quantiferon‐TB were ordered. The urine was normal and a urinary antigen test for Legionella was negative. Contrast‐enhanced computed tomography scan of the chest and abdomen was normal except for mild splenomegaly and an enlarged left axillary lymph node.

The ferritin may have been ordered to help evaluate for Still's disease, which is characterized by sustained fever, lymphadenopathy, and transient rash; however, the characteristic leukocytosis and arthralgias are absent. The computed tomography findings are most notable for the absence of generalized lymphadenopathy or significant hepatosplenomegaly that is seen in lymphoma, leukemia, and lymphotropic processes such as acute EBV infection. The localization of disease to the skin (where the predominant lymphocytes are of T cell origin) with relatively modest lymphadenopathy suggests a T cell lymphoma, perhaps of an indolent variety. Vertical transmission of HTLV‐1 decades ago would make adult T cell leukemia or lymphoma a major consideration.

On the third hospital day, WBC count was 1800/L with 67% neutrophils, 22% lymphocytes, and 1% atypical lymphocytes; LDH rose to 623 IU/L. He had continued fatigue and high fever while the rash gradually faded with oral antihistamines and steroid ointment. On hospital day 4, bone marrow biopsy and skin biopsy of his left thigh were performed.

The further decline in WBC and rise in LDH are modest and therefore do not significantly modify the differential diagnosis. Likewise, 1% atypical lymphocytosis is too low to pinpoint an etiology. Because unremitting fevers start to extend into their third week without a clear source of infection, the probability of malignancy and autoimmunity rise. Improvement with oral antihistamines and topical steroids frequently suggests an underlying allergic process, but the remainder of the clinical picture is not in keeping with atopy or allergy. Cutaneous lymphomas (eg, mycosis fungoides) can have waxing and waning skin manifestations, and can be temporarily or definitively treated by topical steroids. The persistence of his fatigue is of concern given the absence of anemia, cardiopulmonary involvement, or motor weakness.

Bone marrow biopsy showed normocellular marrow with no abnormal cells and some activated macrophages with hemophagocytic activity. Skin biopsy failed to show specific pathology.

His left cervical lymph nodes gradually enlarged. Ultrasound of the neck showed multiple enlarged lymph nodes (left side dominant) with dimension of 17 mm 9 mm 31 mm. Blood and urine cultures returned negative, as did HIV antibody. cytomegalovirus and EBV serologies were consistent with previous infection and the ferritin level was 578 ng/mL (normal, 39‐340 ng/mL). Toxoplasma serology and HTLV‐1 antibody were ordered.

The absence of malignant cells on bone marrow biopsy does not exclude lymphoma, but makes a myelophthisic cause of the cytopenias less likely. The macrophage hemophagocytosis reflects immune activation, which in turn is usually caused by the same viral infections, autoimmune conditions, and lymphoproliferative disorders which constitute the current differential diagnosis.

Bone marrow and skin biopsies are both subject to sampling error, and detection of cutaneous T cell lymphoma is notoriously difficult. However, taken together, the absence of cancer on 2 specimens reduces that possibility.

Sustained unilateral cervical lymphadenopathy with fever in a young Japanese man without any histologic evidence of lymphoma points to Kikuchi's disease, ie, lymphadenitis of unknown etiology associated with varying degrees of systemic manifestations. Fever is a frequent feature, we believe, but diffuse sustained rash, cytopenias, and headache are less common or are seen in severe forms of the disease. The diagnosis of Kikuchi's requires the diligent exclusion of SLE and lymphoma. Examination of the peripheral smear and a lymph node biopsy are required.

Of note, there is also a localized form of Castleman's disease, a nonmalignant lymphoproliferative disorder, that similarly is characterized by focal lymphadenopathy. In distinction to Kikuchi's, however, localized Castleman's is largely asymptomatic and responds marvelously to excision.

On hospital day 9, an excisional biopsy of his left anterior cervical lymph nodes was performed, which revealed paracortical foci with necrosis and a histiocytic cellular infiltrate consistent with subacute necrotizing lymphadenitis (Kikuchi‐Fujimoto disease). Antinuclear antibody, Toxoplasma, and HTLV‐1 antibodies returned negative.

There is no treatment for Kikuchi's. It is usually self‐limited, but steroids are sometimes given for symptomatic control.

His condition began to improve after hospital day 9 without specific treatment, including his WBC count and LDH level. He was discharged home on hospital day 15. In the outpatient clinic 1 and 3 months later, he was well and active without recurrences of any symptoms or laboratory abnormalities. His WBC count was 6600/L and LDH was 268 IU/L.

Commentary

Kikuchi‐Fujimoto disease (KFD), also called Kikuchi's disease, is a benign histiocytic necrotizing lymphadenitis described by both Kikuchi and Fujimoto in 1972.1, 2 It is rare in the United States, but seems more common in Asia, especially Japan, where at least 143 cases have been reported since 1972. The etiology has not been determined, but a viral causeincluding EBV, and human herpesvirus 6 and 8has been suggested.3 An autoimmune etiology is also implicated because of infrequent association with SLE. In general, young women are most likely to be affected. In a review of 244 cases by Kucukardali and colleagues, 77% of patients were female and the mean age was 25; 70% were younger than 30 years of age.4

The common presentation is low‐grade fever with unilateral cervical lymphadenopathy.4 Although generalized lymphadenopathy can occur, it is rare. Other common clinical manifestations include malaise, joint pain, rash, arthritis, and hepatosplenomegaly. No specific laboratory tests for diagnosis are available, but leukopenia (seen in 43% of patients), increased erythrocyte sedimentation rate (40%), and anemia (23%) may be observed.4 In this case, atypical lymphocytes were seen, and are reported in one‐third of patients.5 KFD is generally diagnosed by lymph node biopsy, which typically shows irregular paracortical areas of coagulation necrosis that can distort the nodal architecture, while different types of histiocytes are observed at the margin of necrotic areas.

Other diseases in the differential diagnosisseveral of which were considered by the discussantinclude lymphoma, tuberculosis, SLE, and even metastatic adenocarcinoma. KFD is self‐limited; symptoms typically resolve within 1 to 4 months. Patients with severe manifestations have been treated with anti‐inflammatory drugs and glucocorticosteroids. A recurrence rate of 3% to 4% has been reported.6

The clinicians taking care of this patient initially focused on ruling out those infections occasionally resulting in prolonged fever in a previously healthy young man, such as viruses from the herpes family, HIV, viral hepatitis, tuberculosis, syphilis, infective endocarditis, and intra‐abdominal abscess. Physical examination, specifically lymphadenopathy and mild splenomegaly, made Herpesviridae infections, tuberculosis, syphilis, and lymphoma difficult to exclude. Once the initial evaluation ruled out common infections, attention focused on malignancy and histiocytic necrotizing lymphadenitis, given his ethnicity and geographic location.

The discussant was similarly concerned about infection, malignancy, and noninfectious inflammatory diseases, such as SLE, as possible causes. As evidence of these treatable diseases failed to accumulate, the discussant, an American physician with teaching and clinical experience in Japan, considered endemic diseases such as Behcet's, HTLV‐1, and KFD because they fit the unfolding pattern. Given our global society, clinicians will increasingly benefit from becoming familiar with the less common diseases that afflict the various populations around the world.

In 2003 the Accreditation Council for Graduate Medical Education (ACGME) prescribed residency reform in the form of work hour restrictions without prescribing alternatives to resident based care.1 As a response, many academic medical centers have developed innovative models for providing inpatient care, some of which incorporate Physician Assistants (PAs).2 With further restrictions in resident work hours possible,3 teaching hospitals may increase use of these alternate models to provide inpatient care. Widespread implementation of such new and untested models could impact the care of the approximately 20 million hospitalizations that occur every year in US teaching hospitals.4

Few reports have compared the care delivered by these alternate models with the care provided by traditional resident‐based models of care.58 Roy et al.8 have provided the only recent comparison of a PA‐based model of care with a resident‐based model. They showed lower adjusted costs of inpatient care associated with PA based care but other outcomes were similar to resident‐based teams.

The objective of this study is to provide a valid and usable comparison of the outcomes of a hospitalist‐PA (H‐PA) model of inpatient care with the traditional resident‐based model. This will add to the quantity and quality of the limited research on PA‐based inpatient care, and informs the anticipated increase in the involvement of PAs in this arena.

Methods

Results

Study Population

Of the 52,391 hospitalizations to our hospital during the study period, 13,058 were admitted to general medicine. We excluded 3102 weekend admissions and 209 who met other exclusion criteria. We could not determine the team assignment for 66. Of the remaining 9681 hospitalizations, we assigned 2171 to H‐PA teams and 7510 to RES teams (Figure 1).

Figure 1

Interactions

On adding interaction terms between the team assignment and the fixed effect variables in each model we detected that the effect of H‐PA care on LOS (P < 0.001) and charges (P < 0.001) varied by time of admission (Figure 2a and b). Hospitalizations to H‐PA teams from 6.00 _PM to 6.00 _AM had greater relative increases in LOS as compared to hospitalizations to RES teams during those times. Similarly, hospitalizations during the period 3.00 _PM to 3.00 _AM had relatively higher charges associated with H‐PA care compared to RES care.

Figure 2

Discussion

We found that hospitalizations to our H‐PA teams had longer LOS but similar charges, readmission rates, and mortality as compared to traditional resident‐based teams. These findings were robust to multiple sensitivity and subgroup analyses but when we examined times when both types of teams could receive admissions, the difference in LOS was markedly attenuated and nonsignificant.

We note that most prior reports comparing PA‐based models of inpatient care predate the ACGME work hour regulations. In a randomized control trial (1987‐1988) Simmer et al.5 showed lower lengths of stay and charges but possibly higher risk of readmission for PA based teams as compared to resident based teams. Van Rhee et al.7 conducted a nonrandomized retrospective cohort study (1994‐1995) using administrative data which showed lower resource utilization for PA‐based inpatient care. Our results from 2005 to 2006 reflect the important changes in the organization and delivery of inpatient care since these previous investigations.

Roy et al.8 report the only previously published comparison of PA and resident based GM inpatient care after the ACGME mandated work hour regulations. They found PA‐based care was associated with lower costs, whereas we found similar charges for admissions to RES and H‐PA teams. They also found that LOS was similar for PA and resident‐based care, while we found a higher LOS for admissions to our H‐PA team. We note that although the design of Roy's study was similar to our own, patients cared for by PA‐based teams were geographically localized in their model. This may contribute to the differences in results noted between our studies.

Despite no designed differences in patients assigned to either type of team other than time of admission we noted some differences between the H‐PA and RES teams in the descriptive analysis. These differences, such as a higher proportion of hospitalizations to H‐PA teams being admitted from the ER, residing on nonmedicine wards or having an ICU stay are likely a result of our system of assigning admissions to H‐PA teams early during the workday. For example patients on H‐PA teams were more often located on nonmedicine wards as a result of later discharges and bed availability on medicine wards. The difference that deserves special comment is the much higher proportion (13.8% vs. 6.7%) of hospitalizations with an ICU stay on the H‐PA teams. Hospitalizations directly to the ICU were excluded from our study which means that the hospitalizations with an ICU stay in our study were initially admitted to either H‐PA or RES teams and then transferred to the ICU. Transfers out of the ICU usually occur early in the workday when H‐PA teams accepted patients per our admission schedule. These patients may have been preferentially assigned to H‐PA teams, if on returning from the ICU the original team's resident had changed (and the bounce back rule was not in effect). Importantly, the conclusions of our research are not altered on controlling for this difference in the teams by excluding hospitalizations with an ICU stay.

Hospitalizations to H‐PA teams were associated with higher resource utilization if they occurred later in the day or overnight (Figure 2a and b). During these times a transition of care occurred shortly after admission. For a late day admission the H‐PA teams would transfer care for overnight cross cover soon after the admission and for patients admitted overnight as overflow they would assume care of a patient from the nighttime covering physician performing the admission. On the other hand, on RES teams, interns admitting patients overnight continued to care for their patients for part of the following day (30‐hour call). Similar findings of higher resource utilization associated with transfer of care after admission in the daytime11 and nighttime12 have been previously reported. An alternative hypothesis for our findings is that the hospital maybe busier and thus less efficient during times when H‐PA teams had to admit later in the day or accept patients admitted overnight as overflow. Future research to determine the cause of this significant interaction between team assignment and time of admission on resource utilization is important as the large increases in LOS (up to 30%) and charges (up to 50%) noted, could have a potentially large impact if a higher proportion of hospitalizations were affected by this phenomenon.

Our H‐PA teams were assigned equally complex patients as our RES teams, in contrast to previous reports.8, 13 This was accomplished while improving the resident's educational experience and we have previously reported increases in our resident's board pass rates and in‐service training exam scores with that introduction of our H‐PA teams.14 We thus believe that selection of less complex patients to H‐PA teams such as ours is unnecessary and may give them a second tier status in academic settings.

Our report has limitations. It is a retrospective, nonrandomized investigation using a single institution's administrative database and has the limitations of not being able to account for unmeasured confounders, severity of illness, errors in the database, selection bias and has limited generalizability. We measured charges not actual costs,15 but we feel charges are a true reflection of relative resource use when compared between similar patients within a single institution. We also did not account for the readmissions that occur to other hospitals16 and our results do not reflect resource utilization for the healthcare system in total. For example, we could not tell if higher LOS on H‐PA teams resulted in lower readmissions for their patients in all hospitals in the region, which may reveal an overall resource savings. Additionally, we measured in‐hospital mortality and could not capture deaths related to hospital care that may occur shortly after discharge.

ACGME has proposed revised standards that may further restrict resident duty hours when they take effect in July 2011.3 This may lead to further decreases in resident‐based inpatient care. Teaching hospitals will need to continue to develop alternate models for inpatient care that do not depend on house staff. Our findings provide important evidence to inform the development of such models. Our study shows that one such model: PAs paired with hospitalists, accepting admissions early in the workday, with hospitalist coverage over the weekend and nights can care for GM inpatients as complex as those cared for by resident‐based teams without increasing readmission rates, inpatient mortality, or charges but at the cost of slightly higher LOS.

In 2003 the Accreditation Council for Graduate Medical Education (ACGME) prescribed residency reform in the form of work hour restrictions without prescribing alternatives to resident based care.1 As a response, many academic medical centers have developed innovative models for providing inpatient care, some of which incorporate Physician Assistants (PAs).2 With further restrictions in resident work hours possible,3 teaching hospitals may increase use of these alternate models to provide inpatient care. Widespread implementation of such new and untested models could impact the care of the approximately 20 million hospitalizations that occur every year in US teaching hospitals.4

Few reports have compared the care delivered by these alternate models with the care provided by traditional resident‐based models of care.58 Roy et al.8 have provided the only recent comparison of a PA‐based model of care with a resident‐based model. They showed lower adjusted costs of inpatient care associated with PA based care but other outcomes were similar to resident‐based teams.

The objective of this study is to provide a valid and usable comparison of the outcomes of a hospitalist‐PA (H‐PA) model of inpatient care with the traditional resident‐based model. This will add to the quantity and quality of the limited research on PA‐based inpatient care, and informs the anticipated increase in the involvement of PAs in this arena.

Methods

Results

Study Population

Of the 52,391 hospitalizations to our hospital during the study period, 13,058 were admitted to general medicine. We excluded 3102 weekend admissions and 209 who met other exclusion criteria. We could not determine the team assignment for 66. Of the remaining 9681 hospitalizations, we assigned 2171 to H‐PA teams and 7510 to RES teams (Figure 1).

Figure 1

Interactions

On adding interaction terms between the team assignment and the fixed effect variables in each model we detected that the effect of H‐PA care on LOS (P < 0.001) and charges (P < 0.001) varied by time of admission (Figure 2a and b). Hospitalizations to H‐PA teams from 6.00 _PM to 6.00 _AM had greater relative increases in LOS as compared to hospitalizations to RES teams during those times. Similarly, hospitalizations during the period 3.00 _PM to 3.00 _AM had relatively higher charges associated with H‐PA care compared to RES care.

Figure 2

Discussion

We found that hospitalizations to our H‐PA teams had longer LOS but similar charges, readmission rates, and mortality as compared to traditional resident‐based teams. These findings were robust to multiple sensitivity and subgroup analyses but when we examined times when both types of teams could receive admissions, the difference in LOS was markedly attenuated and nonsignificant.

We note that most prior reports comparing PA‐based models of inpatient care predate the ACGME work hour regulations. In a randomized control trial (1987‐1988) Simmer et al.5 showed lower lengths of stay and charges but possibly higher risk of readmission for PA based teams as compared to resident based teams. Van Rhee et al.7 conducted a nonrandomized retrospective cohort study (1994‐1995) using administrative data which showed lower resource utilization for PA‐based inpatient care. Our results from 2005 to 2006 reflect the important changes in the organization and delivery of inpatient care since these previous investigations.

Roy et al.8 report the only previously published comparison of PA and resident based GM inpatient care after the ACGME mandated work hour regulations. They found PA‐based care was associated with lower costs, whereas we found similar charges for admissions to RES and H‐PA teams. They also found that LOS was similar for PA and resident‐based care, while we found a higher LOS for admissions to our H‐PA team. We note that although the design of Roy's study was similar to our own, patients cared for by PA‐based teams were geographically localized in their model. This may contribute to the differences in results noted between our studies.

Despite no designed differences in patients assigned to either type of team other than time of admission we noted some differences between the H‐PA and RES teams in the descriptive analysis. These differences, such as a higher proportion of hospitalizations to H‐PA teams being admitted from the ER, residing on nonmedicine wards or having an ICU stay are likely a result of our system of assigning admissions to H‐PA teams early during the workday. For example patients on H‐PA teams were more often located on nonmedicine wards as a result of later discharges and bed availability on medicine wards. The difference that deserves special comment is the much higher proportion (13.8% vs. 6.7%) of hospitalizations with an ICU stay on the H‐PA teams. Hospitalizations directly to the ICU were excluded from our study which means that the hospitalizations with an ICU stay in our study were initially admitted to either H‐PA or RES teams and then transferred to the ICU. Transfers out of the ICU usually occur early in the workday when H‐PA teams accepted patients per our admission schedule. These patients may have been preferentially assigned to H‐PA teams, if on returning from the ICU the original team's resident had changed (and the bounce back rule was not in effect). Importantly, the conclusions of our research are not altered on controlling for this difference in the teams by excluding hospitalizations with an ICU stay.

Hospitalizations to H‐PA teams were associated with higher resource utilization if they occurred later in the day or overnight (Figure 2a and b). During these times a transition of care occurred shortly after admission. For a late day admission the H‐PA teams would transfer care for overnight cross cover soon after the admission and for patients admitted overnight as overflow they would assume care of a patient from the nighttime covering physician performing the admission. On the other hand, on RES teams, interns admitting patients overnight continued to care for their patients for part of the following day (30‐hour call). Similar findings of higher resource utilization associated with transfer of care after admission in the daytime11 and nighttime12 have been previously reported. An alternative hypothesis for our findings is that the hospital maybe busier and thus less efficient during times when H‐PA teams had to admit later in the day or accept patients admitted overnight as overflow. Future research to determine the cause of this significant interaction between team assignment and time of admission on resource utilization is important as the large increases in LOS (up to 30%) and charges (up to 50%) noted, could have a potentially large impact if a higher proportion of hospitalizations were affected by this phenomenon.

Our H‐PA teams were assigned equally complex patients as our RES teams, in contrast to previous reports.8, 13 This was accomplished while improving the resident's educational experience and we have previously reported increases in our resident's board pass rates and in‐service training exam scores with that introduction of our H‐PA teams.14 We thus believe that selection of less complex patients to H‐PA teams such as ours is unnecessary and may give them a second tier status in academic settings.

Our report has limitations. It is a retrospective, nonrandomized investigation using a single institution's administrative database and has the limitations of not being able to account for unmeasured confounders, severity of illness, errors in the database, selection bias and has limited generalizability. We measured charges not actual costs,15 but we feel charges are a true reflection of relative resource use when compared between similar patients within a single institution. We also did not account for the readmissions that occur to other hospitals16 and our results do not reflect resource utilization for the healthcare system in total. For example, we could not tell if higher LOS on H‐PA teams resulted in lower readmissions for their patients in all hospitals in the region, which may reveal an overall resource savings. Additionally, we measured in‐hospital mortality and could not capture deaths related to hospital care that may occur shortly after discharge.

ACGME has proposed revised standards that may further restrict resident duty hours when they take effect in July 2011.3 This may lead to further decreases in resident‐based inpatient care. Teaching hospitals will need to continue to develop alternate models for inpatient care that do not depend on house staff. Our findings provide important evidence to inform the development of such models. Our study shows that one such model: PAs paired with hospitalists, accepting admissions early in the workday, with hospitalist coverage over the weekend and nights can care for GM inpatients as complex as those cared for by resident‐based teams without increasing readmission rates, inpatient mortality, or charges but at the cost of slightly higher LOS.

In 2003 the Accreditation Council for Graduate Medical Education (ACGME) prescribed residency reform in the form of work hour restrictions without prescribing alternatives to resident based care.1 As a response, many academic medical centers have developed innovative models for providing inpatient care, some of which incorporate Physician Assistants (PAs).2 With further restrictions in resident work hours possible,3 teaching hospitals may increase use of these alternate models to provide inpatient care. Widespread implementation of such new and untested models could impact the care of the approximately 20 million hospitalizations that occur every year in US teaching hospitals.4

Few reports have compared the care delivered by these alternate models with the care provided by traditional resident‐based models of care.58 Roy et al.8 have provided the only recent comparison of a PA‐based model of care with a resident‐based model. They showed lower adjusted costs of inpatient care associated with PA based care but other outcomes were similar to resident‐based teams.

The objective of this study is to provide a valid and usable comparison of the outcomes of a hospitalist‐PA (H‐PA) model of inpatient care with the traditional resident‐based model. This will add to the quantity and quality of the limited research on PA‐based inpatient care, and informs the anticipated increase in the involvement of PAs in this arena.

Methods

Results

Study Population

Of the 52,391 hospitalizations to our hospital during the study period, 13,058 were admitted to general medicine. We excluded 3102 weekend admissions and 209 who met other exclusion criteria. We could not determine the team assignment for 66. Of the remaining 9681 hospitalizations, we assigned 2171 to H‐PA teams and 7510 to RES teams (Figure 1).

Figure 1

Interactions

On adding interaction terms between the team assignment and the fixed effect variables in each model we detected that the effect of H‐PA care on LOS (P < 0.001) and charges (P < 0.001) varied by time of admission (Figure 2a and b). Hospitalizations to H‐PA teams from 6.00 _PM to 6.00 _AM had greater relative increases in LOS as compared to hospitalizations to RES teams during those times. Similarly, hospitalizations during the period 3.00 _PM to 3.00 _AM had relatively higher charges associated with H‐PA care compared to RES care.

Figure 2

Discussion

We found that hospitalizations to our H‐PA teams had longer LOS but similar charges, readmission rates, and mortality as compared to traditional resident‐based teams. These findings were robust to multiple sensitivity and subgroup analyses but when we examined times when both types of teams could receive admissions, the difference in LOS was markedly attenuated and nonsignificant.

We note that most prior reports comparing PA‐based models of inpatient care predate the ACGME work hour regulations. In a randomized control trial (1987‐1988) Simmer et al.5 showed lower lengths of stay and charges but possibly higher risk of readmission for PA based teams as compared to resident based teams. Van Rhee et al.7 conducted a nonrandomized retrospective cohort study (1994‐1995) using administrative data which showed lower resource utilization for PA‐based inpatient care. Our results from 2005 to 2006 reflect the important changes in the organization and delivery of inpatient care since these previous investigations.

Roy et al.8 report the only previously published comparison of PA and resident based GM inpatient care after the ACGME mandated work hour regulations. They found PA‐based care was associated with lower costs, whereas we found similar charges for admissions to RES and H‐PA teams. They also found that LOS was similar for PA and resident‐based care, while we found a higher LOS for admissions to our H‐PA team. We note that although the design of Roy's study was similar to our own, patients cared for by PA‐based teams were geographically localized in their model. This may contribute to the differences in results noted between our studies.

Despite no designed differences in patients assigned to either type of team other than time of admission we noted some differences between the H‐PA and RES teams in the descriptive analysis. These differences, such as a higher proportion of hospitalizations to H‐PA teams being admitted from the ER, residing on nonmedicine wards or having an ICU stay are likely a result of our system of assigning admissions to H‐PA teams early during the workday. For example patients on H‐PA teams were more often located on nonmedicine wards as a result of later discharges and bed availability on medicine wards. The difference that deserves special comment is the much higher proportion (13.8% vs. 6.7%) of hospitalizations with an ICU stay on the H‐PA teams. Hospitalizations directly to the ICU were excluded from our study which means that the hospitalizations with an ICU stay in our study were initially admitted to either H‐PA or RES teams and then transferred to the ICU. Transfers out of the ICU usually occur early in the workday when H‐PA teams accepted patients per our admission schedule. These patients may have been preferentially assigned to H‐PA teams, if on returning from the ICU the original team's resident had changed (and the bounce back rule was not in effect). Importantly, the conclusions of our research are not altered on controlling for this difference in the teams by excluding hospitalizations with an ICU stay.

Hospitalizations to H‐PA teams were associated with higher resource utilization if they occurred later in the day or overnight (Figure 2a and b). During these times a transition of care occurred shortly after admission. For a late day admission the H‐PA teams would transfer care for overnight cross cover soon after the admission and for patients admitted overnight as overflow they would assume care of a patient from the nighttime covering physician performing the admission. On the other hand, on RES teams, interns admitting patients overnight continued to care for their patients for part of the following day (30‐hour call). Similar findings of higher resource utilization associated with transfer of care after admission in the daytime11 and nighttime12 have been previously reported. An alternative hypothesis for our findings is that the hospital maybe busier and thus less efficient during times when H‐PA teams had to admit later in the day or accept patients admitted overnight as overflow. Future research to determine the cause of this significant interaction between team assignment and time of admission on resource utilization is important as the large increases in LOS (up to 30%) and charges (up to 50%) noted, could have a potentially large impact if a higher proportion of hospitalizations were affected by this phenomenon.

Our H‐PA teams were assigned equally complex patients as our RES teams, in contrast to previous reports.8, 13 This was accomplished while improving the resident's educational experience and we have previously reported increases in our resident's board pass rates and in‐service training exam scores with that introduction of our H‐PA teams.14 We thus believe that selection of less complex patients to H‐PA teams such as ours is unnecessary and may give them a second tier status in academic settings.

Our report has limitations. It is a retrospective, nonrandomized investigation using a single institution's administrative database and has the limitations of not being able to account for unmeasured confounders, severity of illness, errors in the database, selection bias and has limited generalizability. We measured charges not actual costs,15 but we feel charges are a true reflection of relative resource use when compared between similar patients within a single institution. We also did not account for the readmissions that occur to other hospitals16 and our results do not reflect resource utilization for the healthcare system in total. For example, we could not tell if higher LOS on H‐PA teams resulted in lower readmissions for their patients in all hospitals in the region, which may reveal an overall resource savings. Additionally, we measured in‐hospital mortality and could not capture deaths related to hospital care that may occur shortly after discharge.

ACGME has proposed revised standards that may further restrict resident duty hours when they take effect in July 2011.3 This may lead to further decreases in resident‐based inpatient care. Teaching hospitals will need to continue to develop alternate models for inpatient care that do not depend on house staff. Our findings provide important evidence to inform the development of such models. Our study shows that one such model: PAs paired with hospitalists, accepting admissions early in the workday, with hospitalist coverage over the weekend and nights can care for GM inpatients as complex as those cared for by resident‐based teams without increasing readmission rates, inpatient mortality, or charges but at the cost of slightly higher LOS.

Dual antiplatelet therapy (DAPT) (aspirin plus a thienopyridine: clopidogrel or prasugrel) has become the standard treatment for patients with acute coronary syndromes (ACS) and after coronary stent placement (Table 1). Anticoagulant therapy with warfarin is indicated for stroke prevention in atrial fibrillation (AF), profound left ventricular dysfunction, and after mechanical heart valve replacement, as well as for treatment of deep venous thrombosis and pulmonary embolism (Table 2). It is estimated that 41% of the U.S. population over age 40 years is on some form of antiplatelet therapy,6 and 2.5 million patients, mostly elderly, are on long‐term warfarin therapy.7 More specifically, 5% of patients undergoing percutaneous coronary interventions (PCIs) also have an indication for warfarin.8 With widespread use of drug‐eluting stents (DES), the need for a longer duration of DAPT, and the increased age and complexity of hospitalized patients, the safety and challenges of triple therapy (combined DAPT and warfarin) have become more important to the practice of hospital medicine. Triple therapy may increase hospitalization rates, as the risk of major bleeding is four to five times higher than with DAPT.911 In contrast, DAPT is much less effective than warfarin alone in preventing embolic events in AF,12 and warfarin alone or in combination with aspirin (ASA) is inadequate therapy to prevent stent thrombosis. Even fewer data exist on the efficacy and safety of triple therapy in patients with mechanical valves or left ventricular dysfunction.

Hospitalists commonly care for patients on triple therapy; certain indications are appropriate and supported from the available literature while others lack evidence. Knowledge of existing practice guidelines and of supporting research studies leads to optimal management of these complicated patients, and minimizes excessive morbidity from bleeding complications or thromboembolic events such as strokes and stent thrombosis.

In the first part of this article, we present the evidence that supports current recommendations for DAPT or warfarin in specific medical conditions. We also address controversies and unanswered questions. The second part of this review focuses on the available data and provides guidance on the optimal care of patients on triple therapy.

Dual Antiplatelet Therapy Following Acute Coronary Syndromes

Table 3 summarizes key randomized trials of DAPT versus ASA alone in several clinical scenarios. The addition of clopidogrel to ASA in patients with nonST‐elevation ACS reduced the risk of adverse ischemic outcomes in the clopidogrel in unstable angina to prevent recurrent events (CURE) trial,15 as well as in its substudy, the PCI‐CURE (patients with ACS who have undergone stenting).17 In the main CURE study, the study groups diverged within the first 30 days after randomization and the benefit of DAPT persisted throughout the 12 months of the study period. DAPT is also superior to ASA in patients with ST‐elevation myocardial infarction (MI) (CLARITYTIMI 28 and COMMIT trials).13, 14 On the basis of these findings, DAPT has become the standard of care for patients with ACS. The American College of Cardiology (ACC)/American Heart Association (AHA)3 and the European Society of Cardiology18 recommend ASA treatment indefinitely for patients with ACS whether or not they underwent PCI. Clopidogrel is recommended for at least 12 months following ACS, especially for patients who receive a coronary stent.

Despite the proven efficacy of DAPT in ACS, about 15% of patients die or experience reinfarction within 30 days of diagnosis.19 The continued risk for thrombotic events could be due to delayed onset of platelet inhibition and to patient heterogeneity in responsiveness to therapy with ASA and/or clopidogrel.20 Consequently, the optimum dose for clopidogrel and ASA following ACS is uncertain. The CURRENT‐OASIS 7 trial evaluated the efficacy and safety of high‐dose clopidogrel (600‐mg loading dose, 150 mg once daily for 7 days, followed by 75 mg/d) versus standard‐dose clopidogrel (300‐mg loading dose, followed by 75 mg/d) and ASA (75‐100 mg versus 300‐325 mg/d) in patients with ACS who were treated medically, with or without stenting.21 In the overall study population as well as in patients who did not receive stenting, there was no significant difference in the combined rate of death from cardiovascular causes, MI, and stroke between patients receiving the high‐dose and the standard‐dose clopidogrel (4.2% vs 4.4%; P = .37) and high‐dose versus low‐dose ASA (4.2% vs 4.4%; P = .47). There were no significant differences in bleeding complications between the two clopidogrel treatment arms or between the high‐dose and low‐dose ASA groups.

The ACC/AHA guidelines recommend ASA, 75‐162 mg/d indefinitely after medical therapy without stenting (class I, level of evidence: A)3 and clopidogrel 75 mg/d for at least 1 month (class IA) and optimally for 1 year (class IB). Clopidogrel monotherapy is appropriate for patients with ACS who are unable to tolerate ASA due to either hypersensitivity or recent significant gastrointestinal bleeding.

As is the case after coronary stenting, interruption of DAPT soon after ACS may subject patients to high recurrence of cardiovascular events, although few data are available to support this observation. Interruption of DAPT due to bleeding complications or surgical procedures more than 1 month after ACS may be reasonable for a patient who did not receive a stent. Clinicians should restart DAPT after the surgical procedure once the bleeding risk becomes acceptable.

Dual Antiplatelet Therapy Following Coronary Stenting

Warfarin After Acute Coronary Syndromes

Warfarin with different international normalized ratio (INR) goals alone or in combination with ASA has been evaluated after ACS. In an early trial, patients with recent (mean interval 27 days) MI were treated with warfarin alone versus placebo.41 Warfarin conferred a relative risk reduction in mortality of 24% (95% CI, 4‐44%; P = .027) at the expense of major bleeding rates of 0.6%/y. In the ASPECT trial,42 moderate to high intensity anticoagulation after MI resulted in a 53% and 40% reduction in the relative risk of reinfarction (annual incidence 2.3% versus 5.1%) and cerebrovascular events (annual incidence 0.7% versus 1.2%), respectively. In the WARIS II43 and ASPECT‐244 trials, moderate intensity warfarin (INR 2.0‐2.5) in combination with low‐dose ASA, compared with ASA alone, reduced the composite occurrence of death or nonfatal reinfarction, as well as recurrent coronary occlusion after ST‐segment elevation MI. High‐intensity warfarin therapy alone (INR 3.0‐4.0 for ASPECT, 2.8‐4.2 for WARISII) reduced ischemic vascular events compared with ASA alone. Not unexpectedly, major bleeding episodes were more common among patients receiving warfarin.

No randomized trials have compared DAPT with warfarin plus ASA for patients with ACS who did not receive stents. The ACC/AHA guidelines recommend warfarin for secondary prevention following ACS (class IIb). High‐intensity warfarin alone (INR 2.5‐3.5) or moderate intensity (INR 2.0‐2.5) with low‐dose ASA (75‐81 mg/d) may be reasonable for patients at high ischemic and low bleeding risk who are intolerant of clopidogrel (level of evidence: B). Fixed dose warfarin is not recommended by the ACC/AHA primarily on the basis of the Coumadin Aspirin Reinfarction Study (CARS) results. This study of patients following MI was discontinued prematurely because of a lack of incremental benefit of reduced‐dose ASA (80 mg/d) combined with either 1 or 3 mg of warfarin daily when compared with 160 mg/d of ASA alone.

Triple Therapy for PCI and Atrial Fibrillation

AF is the most frequent indication (70%) for long‐term therapy with warfarin in patients scheduled for stent placement.10 Clinical trials have shown that warfarin alone is superior to ASA, clopidogrel, or DAPT for prevention of stroke in patients with AF.45, 46 Although warfarin is indispensable in these settings, DAPT is similarly necessary after stent implantation. As triple therapy increases the risk of bleeding, the management of patients with AF and who have received stents remains controversial. This situation is particularly problematic among patients who have received DES and may benefit from extended DAPT. No randomized trials exist to clarify the optimal treatment in these patients; and the feasibility of such studies is questionable. Small, mostly retrospective, studies (Table 4) provide limited guidance on this issue; most studies focus on bleeding events rather than the cardiovascular efficacy of triple therapy. Because of these limitations, cardiovascular societies give IIb recommendation for either triple therapy or the combination of warfarin and clopidogrel in this setting and the level of evidence is C.1, 59, 60

In the largest study to date, Nguyen et al.58 evaluated 800 patients who underwent stenting for ACS and were discharged on warfarin plus single antiplatelet agent or triple therapy as part of the GRACE registry. At 6 months, triple therapy conferred a significant reduction in stroke (0.7% versus 3.4%, P = .02) but not in death or MI. There were no differences in in‐hospital major bleeding events between the two groups (5.9% versus 4.6%; P = .46). Similarly, Sarafoff et al.53 reported no significant differences in the combined endpoint (death, MI, stent thrombosis or stroke) or bleeding complications among patients who received triple therapy or DAPT at 2 years of follow‐up. In contrast, Ruiz‐Nodar et al.52 showed that triple therapy, compared with DAPT, at discharge reduced the incidence of death (17.8% versus 27.8%; adjusted HR = 3.43; 95% CI, 1.617.54; P = .002) and major adverse cardiac events (26.5% versus 38.7%; adjusted HR = 4.9; 95% CI, 2.1711.1; P = .01), without a substantial increase in major bleeding events.

The value of combination antiplatelet therapy to prevent stent thrombosis in these patients is clearer in the study reported by Karjalainen et al.10 This case‐control study of 239 patients receiving warfarin at baseline who underwent PCI evaluated a primary endpoint of death, MI, target‐vessel revascularization, or stent thrombosis and a secondary endpoint of major bleeding and stroke to 12 months of follow‐up. Forty‐eight percent of patients received triple therapy, whereas 15.5% were discharged on DAPT. The remaining patients received warfarin plus a single antiplatelet agent. Stent thrombosis occurred more frequently among patients receiving warfarin plus ASA (15.2%) than among those receiving triple therapy (1.9%). As expected, stroke was more frequent in patients treated with DAPT (8.8%) than among those receiving triple therapy (2.8%). Major bleeding was similar between groups. Therapy with warfarin was an independent predictor of both major bleeding and major cardiac events at 1 year. This observation illustrates that the outcome of PCI in patients on chronic warfarin therapy is unsatisfactory irrespective of the antithrombotic combinations used, highlighting the need for better strategies to treat these patients.

Choice of Therapy and Management of Patients Eligible for Triple Therapy

Current guidelines for PCI do not provide guidance for patients with an indication for triple therapy due to a paucity of published evidence. Several ongoing prospective trials aim to address the management of these patients (AFCAS, ISAR‐TRIPLE). Pending further study, clinicians should consider the embolic risk (CHADS2 score), target INR, type of stent, bleeding risk, and duration of treatment when determining the appropriate antiplatelet/anticoagulant combinations. The CHADS2 score (Table 2) stratifies the risk for stroke among patients with AF,4 while the Outpatient Bleeding Risk Index (OBRI) allows estimation of bleeding risk.12, 61 The OBRI considers age > 65 years, prior stroke, prior gastrointestinal bleeding, and any of four comorbidities (recent MI, anemia, diabetes, or renal insufficiency) in order to stratify patients into three risk groups.61 Patients with three to four risk factors have a high risk of bleeding (23% at 3 months and 48% at 12 months) whereas patients with no risk factors have only a 3% risk of bleeding at 12 months. Unfortunately, advanced age and prior stroke appear in both OBRI and CHADS.

For patients with AF who are at high risk for embolic stroke (>3% per year), we recommend triple therapy for the shortest time possible, followed by warfarin and ASA indefinitely. In case of BMS, it is acceptable to shorten triple therapy duration to 1 month. The optimal duration of triple therapy for patients with DES is uncertain; recommended durations range from 3 months to 1 year.62 If the potential consequences of stent thrombosis are high due to a large amount of myocardium at risk, an extended period of triple therapy might be justified. For patients whose stroke risk is lower (CHADS2 score of 0‐1), the risk for bleeding likely outweighs any benefit from stroke prevention. In this instance, it is reasonable to use DAPT with ASA and clopidogrel for 1 month after BMS and 12 months after DES, followed by ASA, with or without warfarin, indefinitely. In a recently published study, patients with AF and a CHADS2 score of 1 had a yearly stroke risk of 1.25% while taking DAPT63; the risk of major bleeding for triple therapy is 6.1% per year.64

For patients who have a high bleeding risk, BMS are the preferred stent type as the duration of triple therapy might be limited to 4 weeks. To our knowledge, no randomized study has evaluated the outcome of patients with BMS compared with DES who also have an indication for warfarin. Because studies have suggested that clopidogrel is more effective than aspirin in preventing stent thrombosis and in reducing death or MI after coronary stenting,40, 65 warfarin and single antiplatelet therapy with clopidogrel might be a reasonable treatment option in patients with high bleeding risk. The WOEST study (NCT00769938), currently recruiting participants, is the first randomized study specifically designed to test this hypothesis.

Since gastrointestinal bleeding accounts for approximately 30‐40% of hemorrhagic events in patients on combined ASA and anticoagulant therapy, an expert consensus document recommended concomitant treatment with proton pump inhibitors (PPIs) to reduce this risk.66 In contrast, the 2009 Focused Updates of the ACC/AHA/SCAI Guidelines did not recommend the use of PPIs with DAPT in the setting of ACS.2 This is because of studies that show inhibition of platelet activation,67 and potential clinical harm,68 when clopidogrel is combined with certain PPIs that inhibit the CYP2C19 enzyme. However, to date there are no convincing randomized clinical trial data documenting an important clinical drug‐drug interaction. The U.S. Food and Drug Administration (FDA) advises that physicians avoid the use of clopidogrel in patients with impaired CYP2C19 function due to known genetic variation or due to concomitant use of drugs that inhibit CYP2C19 activity. More specifically, the FDA recommends avoiding the use of omeprazole and esomeprazole in patients taking clopidogrel.69

In particular, elderly patients have an increased risk of bleeding while receiving triple therapy. In a study of patients over age 65, 2.5% were hospitalized for bleeding in the first year after PCI, and the use of triple therapy was the strongest predictor of bleeding (more than threefold increase).70 One in five patients suffered death or MI at 1 year after hospitalization for bleeding.70 The basis for poor outcomes after hospitalization for bleeding in this population is multifactorial and may be due to the location of bleeding, associated hypercoagulable state, potential adverse impact of blood transfusion, withdrawal of warfarin therapy in patients with AF and PCI, and the premature discontinuation of DAPT. The use of nonsteroidal anti‐inflammatory drugs (NSAIDs) is common among the elderly and conferred a doubling of bleeding risk.70 Limiting the use of NSAID, the use of low‐dose ASA beyond 30 days after stent implantation, greater use of BMS, and maintaining INR at the lowest possible level (INR of 22.5) will reduce the risk for bleeding.57, 71

New Anticoagulants

Due to the high risk for bleeding with warfarin and the challenges inherent in INR monitoring, researchers have developed several novel anticoagulants whose advantages include fixed daily dosing and no need for monitoring. Dabigatran is a direct oral thrombin inhibitor that is already licensed in Europe and Canada for thromboprophylaxis after hip or knee surgery. It has also been studied in patients with AF. In the RE‐LY trial, patients with AF who received dabigatran 110 mg daily had rates of stroke and systemic embolism that were similar to those with warfarin, as well as lower rates of major hemorrhage.72 The randomized ReDEEM trial, reported at the AHA 2009 Scientific Sessions, was aimed at finding a dosage of dabigatran that achieves a good balance between clinical effectiveness and bleeding risk when combined with aspirin and clopidogrel after acute MI. Dosages ranging from 50 mg twice daily to 150 mg twice daily were all associated with 6‐month rates of bleeding lower than 2%. Hospitalists should view these encouraging results cautiously until the publication of ReDEEM trial results in a peer‐reviewed journal.

A variety of oral Xa antagonists are also being evaluated in patients with AF or ACS. These trials offer insight into triple therapy regimens that include ASA, clopidogrel, and an Xa antagonist. In a recent study of the oral Xa antagonist rivaroxaban, investigators stratified 3491 subjects with ACS according to whether they received concomitant ASA alone or ASA and clopidogrel.73 Subjects receiving ASA plus rivaroxaban had a modest increase in bleeding. Triple therapy, however, increased the composite bleeding rate from 3.5% in the DAPT group to approximately 6‐15% (low‐dose or high‐dose rivaroxaban, respectively). Rivaroxaban is currently under review by the FDA.

These novel agents might eventually replace warfarin for many or most indications for anticoagulation. It is imperative that future research compare the efficacy and risk of bleeding between triple therapy using these new agents and triple therapy with warfarin.

Conclusions

The management of patients on long‐term anticoagulation who require DAPT because of ACS or coronary stenting is challenging. DAPT may safely substitute for warfarin only for patients at low risk for a thromboembolic event (ie, low‐risk AF with low CHADS2 score). Clinicians should not interrupt warfarin in patients at higher risk (ie, intermediate to high‐risk AF, mechanical valves, or recent venous thromboembolism), even in the presence of DAPT. In these patients, triple therapy is the optimal approach following coronary stenting (and possibly during the initial period after ACS without stenting). As this approach confers a fivefold increase in bleeding complications compared with DAPT, careful monitoring of the INR, the addition of PPIs, and the exclusion of elderly patients who are at the highest risk for bleeding complications74 is recommended. The preferred duration of triple therapy after BMS in patients who require long‐term anticoagulation is 1 month, whereas the optimal duration after ACS or DES remains unresolved.

Dual antiplatelet therapy (DAPT) (aspirin plus a thienopyridine: clopidogrel or prasugrel) has become the standard treatment for patients with acute coronary syndromes (ACS) and after coronary stent placement (Table 1). Anticoagulant therapy with warfarin is indicated for stroke prevention in atrial fibrillation (AF), profound left ventricular dysfunction, and after mechanical heart valve replacement, as well as for treatment of deep venous thrombosis and pulmonary embolism (Table 2). It is estimated that 41% of the U.S. population over age 40 years is on some form of antiplatelet therapy,6 and 2.5 million patients, mostly elderly, are on long‐term warfarin therapy.7 More specifically, 5% of patients undergoing percutaneous coronary interventions (PCIs) also have an indication for warfarin.8 With widespread use of drug‐eluting stents (DES), the need for a longer duration of DAPT, and the increased age and complexity of hospitalized patients, the safety and challenges of triple therapy (combined DAPT and warfarin) have become more important to the practice of hospital medicine. Triple therapy may increase hospitalization rates, as the risk of major bleeding is four to five times higher than with DAPT.911 In contrast, DAPT is much less effective than warfarin alone in preventing embolic events in AF,12 and warfarin alone or in combination with aspirin (ASA) is inadequate therapy to prevent stent thrombosis. Even fewer data exist on the efficacy and safety of triple therapy in patients with mechanical valves or left ventricular dysfunction.

Hospitalists commonly care for patients on triple therapy; certain indications are appropriate and supported from the available literature while others lack evidence. Knowledge of existing practice guidelines and of supporting research studies leads to optimal management of these complicated patients, and minimizes excessive morbidity from bleeding complications or thromboembolic events such as strokes and stent thrombosis.

In the first part of this article, we present the evidence that supports current recommendations for DAPT or warfarin in specific medical conditions. We also address controversies and unanswered questions. The second part of this review focuses on the available data and provides guidance on the optimal care of patients on triple therapy.

Dual Antiplatelet Therapy Following Acute Coronary Syndromes

Table 3 summarizes key randomized trials of DAPT versus ASA alone in several clinical scenarios. The addition of clopidogrel to ASA in patients with nonST‐elevation ACS reduced the risk of adverse ischemic outcomes in the clopidogrel in unstable angina to prevent recurrent events (CURE) trial,15 as well as in its substudy, the PCI‐CURE (patients with ACS who have undergone stenting).17 In the main CURE study, the study groups diverged within the first 30 days after randomization and the benefit of DAPT persisted throughout the 12 months of the study period. DAPT is also superior to ASA in patients with ST‐elevation myocardial infarction (MI) (CLARITYTIMI 28 and COMMIT trials).13, 14 On the basis of these findings, DAPT has become the standard of care for patients with ACS. The American College of Cardiology (ACC)/American Heart Association (AHA)3 and the European Society of Cardiology18 recommend ASA treatment indefinitely for patients with ACS whether or not they underwent PCI. Clopidogrel is recommended for at least 12 months following ACS, especially for patients who receive a coronary stent.

Despite the proven efficacy of DAPT in ACS, about 15% of patients die or experience reinfarction within 30 days of diagnosis.19 The continued risk for thrombotic events could be due to delayed onset of platelet inhibition and to patient heterogeneity in responsiveness to therapy with ASA and/or clopidogrel.20 Consequently, the optimum dose for clopidogrel and ASA following ACS is uncertain. The CURRENT‐OASIS 7 trial evaluated the efficacy and safety of high‐dose clopidogrel (600‐mg loading dose, 150 mg once daily for 7 days, followed by 75 mg/d) versus standard‐dose clopidogrel (300‐mg loading dose, followed by 75 mg/d) and ASA (75‐100 mg versus 300‐325 mg/d) in patients with ACS who were treated medically, with or without stenting.21 In the overall study population as well as in patients who did not receive stenting, there was no significant difference in the combined rate of death from cardiovascular causes, MI, and stroke between patients receiving the high‐dose and the standard‐dose clopidogrel (4.2% vs 4.4%; P = .37) and high‐dose versus low‐dose ASA (4.2% vs 4.4%; P = .47). There were no significant differences in bleeding complications between the two clopidogrel treatment arms or between the high‐dose and low‐dose ASA groups.

The ACC/AHA guidelines recommend ASA, 75‐162 mg/d indefinitely after medical therapy without stenting (class I, level of evidence: A)3 and clopidogrel 75 mg/d for at least 1 month (class IA) and optimally for 1 year (class IB). Clopidogrel monotherapy is appropriate for patients with ACS who are unable to tolerate ASA due to either hypersensitivity or recent significant gastrointestinal bleeding.

As is the case after coronary stenting, interruption of DAPT soon after ACS may subject patients to high recurrence of cardiovascular events, although few data are available to support this observation. Interruption of DAPT due to bleeding complications or surgical procedures more than 1 month after ACS may be reasonable for a patient who did not receive a stent. Clinicians should restart DAPT after the surgical procedure once the bleeding risk becomes acceptable.

Dual Antiplatelet Therapy Following Coronary Stenting

Warfarin After Acute Coronary Syndromes

Warfarin with different international normalized ratio (INR) goals alone or in combination with ASA has been evaluated after ACS. In an early trial, patients with recent (mean interval 27 days) MI were treated with warfarin alone versus placebo.41 Warfarin conferred a relative risk reduction in mortality of 24% (95% CI, 4‐44%; P = .027) at the expense of major bleeding rates of 0.6%/y. In the ASPECT trial,42 moderate to high intensity anticoagulation after MI resulted in a 53% and 40% reduction in the relative risk of reinfarction (annual incidence 2.3% versus 5.1%) and cerebrovascular events (annual incidence 0.7% versus 1.2%), respectively. In the WARIS II43 and ASPECT‐244 trials, moderate intensity warfarin (INR 2.0‐2.5) in combination with low‐dose ASA, compared with ASA alone, reduced the composite occurrence of death or nonfatal reinfarction, as well as recurrent coronary occlusion after ST‐segment elevation MI. High‐intensity warfarin therapy alone (INR 3.0‐4.0 for ASPECT, 2.8‐4.2 for WARISII) reduced ischemic vascular events compared with ASA alone. Not unexpectedly, major bleeding episodes were more common among patients receiving warfarin.

No randomized trials have compared DAPT with warfarin plus ASA for patients with ACS who did not receive stents. The ACC/AHA guidelines recommend warfarin for secondary prevention following ACS (class IIb). High‐intensity warfarin alone (INR 2.5‐3.5) or moderate intensity (INR 2.0‐2.5) with low‐dose ASA (75‐81 mg/d) may be reasonable for patients at high ischemic and low bleeding risk who are intolerant of clopidogrel (level of evidence: B). Fixed dose warfarin is not recommended by the ACC/AHA primarily on the basis of the Coumadin Aspirin Reinfarction Study (CARS) results. This study of patients following MI was discontinued prematurely because of a lack of incremental benefit of reduced‐dose ASA (80 mg/d) combined with either 1 or 3 mg of warfarin daily when compared with 160 mg/d of ASA alone.

Triple Therapy for PCI and Atrial Fibrillation

AF is the most frequent indication (70%) for long‐term therapy with warfarin in patients scheduled for stent placement.10 Clinical trials have shown that warfarin alone is superior to ASA, clopidogrel, or DAPT for prevention of stroke in patients with AF.45, 46 Although warfarin is indispensable in these settings, DAPT is similarly necessary after stent implantation. As triple therapy increases the risk of bleeding, the management of patients with AF and who have received stents remains controversial. This situation is particularly problematic among patients who have received DES and may benefit from extended DAPT. No randomized trials exist to clarify the optimal treatment in these patients; and the feasibility of such studies is questionable. Small, mostly retrospective, studies (Table 4) provide limited guidance on this issue; most studies focus on bleeding events rather than the cardiovascular efficacy of triple therapy. Because of these limitations, cardiovascular societies give IIb recommendation for either triple therapy or the combination of warfarin and clopidogrel in this setting and the level of evidence is C.1, 59, 60

In the largest study to date, Nguyen et al.58 evaluated 800 patients who underwent stenting for ACS and were discharged on warfarin plus single antiplatelet agent or triple therapy as part of the GRACE registry. At 6 months, triple therapy conferred a significant reduction in stroke (0.7% versus 3.4%, P = .02) but not in death or MI. There were no differences in in‐hospital major bleeding events between the two groups (5.9% versus 4.6%; P = .46). Similarly, Sarafoff et al.53 reported no significant differences in the combined endpoint (death, MI, stent thrombosis or stroke) or bleeding complications among patients who received triple therapy or DAPT at 2 years of follow‐up. In contrast, Ruiz‐Nodar et al.52 showed that triple therapy, compared with DAPT, at discharge reduced the incidence of death (17.8% versus 27.8%; adjusted HR = 3.43; 95% CI, 1.617.54; P = .002) and major adverse cardiac events (26.5% versus 38.7%; adjusted HR = 4.9; 95% CI, 2.1711.1; P = .01), without a substantial increase in major bleeding events.

The value of combination antiplatelet therapy to prevent stent thrombosis in these patients is clearer in the study reported by Karjalainen et al.10 This case‐control study of 239 patients receiving warfarin at baseline who underwent PCI evaluated a primary endpoint of death, MI, target‐vessel revascularization, or stent thrombosis and a secondary endpoint of major bleeding and stroke to 12 months of follow‐up. Forty‐eight percent of patients received triple therapy, whereas 15.5% were discharged on DAPT. The remaining patients received warfarin plus a single antiplatelet agent. Stent thrombosis occurred more frequently among patients receiving warfarin plus ASA (15.2%) than among those receiving triple therapy (1.9%). As expected, stroke was more frequent in patients treated with DAPT (8.8%) than among those receiving triple therapy (2.8%). Major bleeding was similar between groups. Therapy with warfarin was an independent predictor of both major bleeding and major cardiac events at 1 year. This observation illustrates that the outcome of PCI in patients on chronic warfarin therapy is unsatisfactory irrespective of the antithrombotic combinations used, highlighting the need for better strategies to treat these patients.

Choice of Therapy and Management of Patients Eligible for Triple Therapy

Current guidelines for PCI do not provide guidance for patients with an indication for triple therapy due to a paucity of published evidence. Several ongoing prospective trials aim to address the management of these patients (AFCAS, ISAR‐TRIPLE). Pending further study, clinicians should consider the embolic risk (CHADS2 score), target INR, type of stent, bleeding risk, and duration of treatment when determining the appropriate antiplatelet/anticoagulant combinations. The CHADS2 score (Table 2) stratifies the risk for stroke among patients with AF,4 while the Outpatient Bleeding Risk Index (OBRI) allows estimation of bleeding risk.12, 61 The OBRI considers age > 65 years, prior stroke, prior gastrointestinal bleeding, and any of four comorbidities (recent MI, anemia, diabetes, or renal insufficiency) in order to stratify patients into three risk groups.61 Patients with three to four risk factors have a high risk of bleeding (23% at 3 months and 48% at 12 months) whereas patients with no risk factors have only a 3% risk of bleeding at 12 months. Unfortunately, advanced age and prior stroke appear in both OBRI and CHADS.

For patients with AF who are at high risk for embolic stroke (>3% per year), we recommend triple therapy for the shortest time possible, followed by warfarin and ASA indefinitely. In case of BMS, it is acceptable to shorten triple therapy duration to 1 month. The optimal duration of triple therapy for patients with DES is uncertain; recommended durations range from 3 months to 1 year.62 If the potential consequences of stent thrombosis are high due to a large amount of myocardium at risk, an extended period of triple therapy might be justified. For patients whose stroke risk is lower (CHADS2 score of 0‐1), the risk for bleeding likely outweighs any benefit from stroke prevention. In this instance, it is reasonable to use DAPT with ASA and clopidogrel for 1 month after BMS and 12 months after DES, followed by ASA, with or without warfarin, indefinitely. In a recently published study, patients with AF and a CHADS2 score of 1 had a yearly stroke risk of 1.25% while taking DAPT63; the risk of major bleeding for triple therapy is 6.1% per year.64

For patients who have a high bleeding risk, BMS are the preferred stent type as the duration of triple therapy might be limited to 4 weeks. To our knowledge, no randomized study has evaluated the outcome of patients with BMS compared with DES who also have an indication for warfarin. Because studies have suggested that clopidogrel is more effective than aspirin in preventing stent thrombosis and in reducing death or MI after coronary stenting,40, 65 warfarin and single antiplatelet therapy with clopidogrel might be a reasonable treatment option in patients with high bleeding risk. The WOEST study (NCT00769938), currently recruiting participants, is the first randomized study specifically designed to test this hypothesis.

Since gastrointestinal bleeding accounts for approximately 30‐40% of hemorrhagic events in patients on combined ASA and anticoagulant therapy, an expert consensus document recommended concomitant treatment with proton pump inhibitors (PPIs) to reduce this risk.66 In contrast, the 2009 Focused Updates of the ACC/AHA/SCAI Guidelines did not recommend the use of PPIs with DAPT in the setting of ACS.2 This is because of studies that show inhibition of platelet activation,67 and potential clinical harm,68 when clopidogrel is combined with certain PPIs that inhibit the CYP2C19 enzyme. However, to date there are no convincing randomized clinical trial data documenting an important clinical drug‐drug interaction. The U.S. Food and Drug Administration (FDA) advises that physicians avoid the use of clopidogrel in patients with impaired CYP2C19 function due to known genetic variation or due to concomitant use of drugs that inhibit CYP2C19 activity. More specifically, the FDA recommends avoiding the use of omeprazole and esomeprazole in patients taking clopidogrel.69

In particular, elderly patients have an increased risk of bleeding while receiving triple therapy. In a study of patients over age 65, 2.5% were hospitalized for bleeding in the first year after PCI, and the use of triple therapy was the strongest predictor of bleeding (more than threefold increase).70 One in five patients suffered death or MI at 1 year after hospitalization for bleeding.70 The basis for poor outcomes after hospitalization for bleeding in this population is multifactorial and may be due to the location of bleeding, associated hypercoagulable state, potential adverse impact of blood transfusion, withdrawal of warfarin therapy in patients with AF and PCI, and the premature discontinuation of DAPT. The use of nonsteroidal anti‐inflammatory drugs (NSAIDs) is common among the elderly and conferred a doubling of bleeding risk.70 Limiting the use of NSAID, the use of low‐dose ASA beyond 30 days after stent implantation, greater use of BMS, and maintaining INR at the lowest possible level (INR of 22.5) will reduce the risk for bleeding.57, 71

New Anticoagulants

Due to the high risk for bleeding with warfarin and the challenges inherent in INR monitoring, researchers have developed several novel anticoagulants whose advantages include fixed daily dosing and no need for monitoring. Dabigatran is a direct oral thrombin inhibitor that is already licensed in Europe and Canada for thromboprophylaxis after hip or knee surgery. It has also been studied in patients with AF. In the RE‐LY trial, patients with AF who received dabigatran 110 mg daily had rates of stroke and systemic embolism that were similar to those with warfarin, as well as lower rates of major hemorrhage.72 The randomized ReDEEM trial, reported at the AHA 2009 Scientific Sessions, was aimed at finding a dosage of dabigatran that achieves a good balance between clinical effectiveness and bleeding risk when combined with aspirin and clopidogrel after acute MI. Dosages ranging from 50 mg twice daily to 150 mg twice daily were all associated with 6‐month rates of bleeding lower than 2%. Hospitalists should view these encouraging results cautiously until the publication of ReDEEM trial results in a peer‐reviewed journal.

A variety of oral Xa antagonists are also being evaluated in patients with AF or ACS. These trials offer insight into triple therapy regimens that include ASA, clopidogrel, and an Xa antagonist. In a recent study of the oral Xa antagonist rivaroxaban, investigators stratified 3491 subjects with ACS according to whether they received concomitant ASA alone or ASA and clopidogrel.73 Subjects receiving ASA plus rivaroxaban had a modest increase in bleeding. Triple therapy, however, increased the composite bleeding rate from 3.5% in the DAPT group to approximately 6‐15% (low‐dose or high‐dose rivaroxaban, respectively). Rivaroxaban is currently under review by the FDA.

These novel agents might eventually replace warfarin for many or most indications for anticoagulation. It is imperative that future research compare the efficacy and risk of bleeding between triple therapy using these new agents and triple therapy with warfarin.

Conclusions

The management of patients on long‐term anticoagulation who require DAPT because of ACS or coronary stenting is challenging. DAPT may safely substitute for warfarin only for patients at low risk for a thromboembolic event (ie, low‐risk AF with low CHADS2 score). Clinicians should not interrupt warfarin in patients at higher risk (ie, intermediate to high‐risk AF, mechanical valves, or recent venous thromboembolism), even in the presence of DAPT. In these patients, triple therapy is the optimal approach following coronary stenting (and possibly during the initial period after ACS without stenting). As this approach confers a fivefold increase in bleeding complications compared with DAPT, careful monitoring of the INR, the addition of PPIs, and the exclusion of elderly patients who are at the highest risk for bleeding complications74 is recommended. The preferred duration of triple therapy after BMS in patients who require long‐term anticoagulation is 1 month, whereas the optimal duration after ACS or DES remains unresolved.

Dual antiplatelet therapy (DAPT) (aspirin plus a thienopyridine: clopidogrel or prasugrel) has become the standard treatment for patients with acute coronary syndromes (ACS) and after coronary stent placement (Table 1). Anticoagulant therapy with warfarin is indicated for stroke prevention in atrial fibrillation (AF), profound left ventricular dysfunction, and after mechanical heart valve replacement, as well as for treatment of deep venous thrombosis and pulmonary embolism (Table 2). It is estimated that 41% of the U.S. population over age 40 years is on some form of antiplatelet therapy,6 and 2.5 million patients, mostly elderly, are on long‐term warfarin therapy.7 More specifically, 5% of patients undergoing percutaneous coronary interventions (PCIs) also have an indication for warfarin.8 With widespread use of drug‐eluting stents (DES), the need for a longer duration of DAPT, and the increased age and complexity of hospitalized patients, the safety and challenges of triple therapy (combined DAPT and warfarin) have become more important to the practice of hospital medicine. Triple therapy may increase hospitalization rates, as the risk of major bleeding is four to five times higher than with DAPT.911 In contrast, DAPT is much less effective than warfarin alone in preventing embolic events in AF,12 and warfarin alone or in combination with aspirin (ASA) is inadequate therapy to prevent stent thrombosis. Even fewer data exist on the efficacy and safety of triple therapy in patients with mechanical valves or left ventricular dysfunction.

Hospitalists commonly care for patients on triple therapy; certain indications are appropriate and supported from the available literature while others lack evidence. Knowledge of existing practice guidelines and of supporting research studies leads to optimal management of these complicated patients, and minimizes excessive morbidity from bleeding complications or thromboembolic events such as strokes and stent thrombosis.

In the first part of this article, we present the evidence that supports current recommendations for DAPT or warfarin in specific medical conditions. We also address controversies and unanswered questions. The second part of this review focuses on the available data and provides guidance on the optimal care of patients on triple therapy.

Dual Antiplatelet Therapy Following Acute Coronary Syndromes

Table 3 summarizes key randomized trials of DAPT versus ASA alone in several clinical scenarios. The addition of clopidogrel to ASA in patients with nonST‐elevation ACS reduced the risk of adverse ischemic outcomes in the clopidogrel in unstable angina to prevent recurrent events (CURE) trial,15 as well as in its substudy, the PCI‐CURE (patients with ACS who have undergone stenting).17 In the main CURE study, the study groups diverged within the first 30 days after randomization and the benefit of DAPT persisted throughout the 12 months of the study period. DAPT is also superior to ASA in patients with ST‐elevation myocardial infarction (MI) (CLARITYTIMI 28 and COMMIT trials).13, 14 On the basis of these findings, DAPT has become the standard of care for patients with ACS. The American College of Cardiology (ACC)/American Heart Association (AHA)3 and the European Society of Cardiology18 recommend ASA treatment indefinitely for patients with ACS whether or not they underwent PCI. Clopidogrel is recommended for at least 12 months following ACS, especially for patients who receive a coronary stent.

Despite the proven efficacy of DAPT in ACS, about 15% of patients die or experience reinfarction within 30 days of diagnosis.19 The continued risk for thrombotic events could be due to delayed onset of platelet inhibition and to patient heterogeneity in responsiveness to therapy with ASA and/or clopidogrel.20 Consequently, the optimum dose for clopidogrel and ASA following ACS is uncertain. The CURRENT‐OASIS 7 trial evaluated the efficacy and safety of high‐dose clopidogrel (600‐mg loading dose, 150 mg once daily for 7 days, followed by 75 mg/d) versus standard‐dose clopidogrel (300‐mg loading dose, followed by 75 mg/d) and ASA (75‐100 mg versus 300‐325 mg/d) in patients with ACS who were treated medically, with or without stenting.21 In the overall study population as well as in patients who did not receive stenting, there was no significant difference in the combined rate of death from cardiovascular causes, MI, and stroke between patients receiving the high‐dose and the standard‐dose clopidogrel (4.2% vs 4.4%; P = .37) and high‐dose versus low‐dose ASA (4.2% vs 4.4%; P = .47). There were no significant differences in bleeding complications between the two clopidogrel treatment arms or between the high‐dose and low‐dose ASA groups.

The ACC/AHA guidelines recommend ASA, 75‐162 mg/d indefinitely after medical therapy without stenting (class I, level of evidence: A)3 and clopidogrel 75 mg/d for at least 1 month (class IA) and optimally for 1 year (class IB). Clopidogrel monotherapy is appropriate for patients with ACS who are unable to tolerate ASA due to either hypersensitivity or recent significant gastrointestinal bleeding.

As is the case after coronary stenting, interruption of DAPT soon after ACS may subject patients to high recurrence of cardiovascular events, although few data are available to support this observation. Interruption of DAPT due to bleeding complications or surgical procedures more than 1 month after ACS may be reasonable for a patient who did not receive a stent. Clinicians should restart DAPT after the surgical procedure once the bleeding risk becomes acceptable.

Dual Antiplatelet Therapy Following Coronary Stenting

Warfarin After Acute Coronary Syndromes

Warfarin with different international normalized ratio (INR) goals alone or in combination with ASA has been evaluated after ACS. In an early trial, patients with recent (mean interval 27 days) MI were treated with warfarin alone versus placebo.41 Warfarin conferred a relative risk reduction in mortality of 24% (95% CI, 4‐44%; P = .027) at the expense of major bleeding rates of 0.6%/y. In the ASPECT trial,42 moderate to high intensity anticoagulation after MI resulted in a 53% and 40% reduction in the relative risk of reinfarction (annual incidence 2.3% versus 5.1%) and cerebrovascular events (annual incidence 0.7% versus 1.2%), respectively. In the WARIS II43 and ASPECT‐244 trials, moderate intensity warfarin (INR 2.0‐2.5) in combination with low‐dose ASA, compared with ASA alone, reduced the composite occurrence of death or nonfatal reinfarction, as well as recurrent coronary occlusion after ST‐segment elevation MI. High‐intensity warfarin therapy alone (INR 3.0‐4.0 for ASPECT, 2.8‐4.2 for WARISII) reduced ischemic vascular events compared with ASA alone. Not unexpectedly, major bleeding episodes were more common among patients receiving warfarin.

No randomized trials have compared DAPT with warfarin plus ASA for patients with ACS who did not receive stents. The ACC/AHA guidelines recommend warfarin for secondary prevention following ACS (class IIb). High‐intensity warfarin alone (INR 2.5‐3.5) or moderate intensity (INR 2.0‐2.5) with low‐dose ASA (75‐81 mg/d) may be reasonable for patients at high ischemic and low bleeding risk who are intolerant of clopidogrel (level of evidence: B). Fixed dose warfarin is not recommended by the ACC/AHA primarily on the basis of the Coumadin Aspirin Reinfarction Study (CARS) results. This study of patients following MI was discontinued prematurely because of a lack of incremental benefit of reduced‐dose ASA (80 mg/d) combined with either 1 or 3 mg of warfarin daily when compared with 160 mg/d of ASA alone.

Triple Therapy for PCI and Atrial Fibrillation

AF is the most frequent indication (70%) for long‐term therapy with warfarin in patients scheduled for stent placement.10 Clinical trials have shown that warfarin alone is superior to ASA, clopidogrel, or DAPT for prevention of stroke in patients with AF.45, 46 Although warfarin is indispensable in these settings, DAPT is similarly necessary after stent implantation. As triple therapy increases the risk of bleeding, the management of patients with AF and who have received stents remains controversial. This situation is particularly problematic among patients who have received DES and may benefit from extended DAPT. No randomized trials exist to clarify the optimal treatment in these patients; and the feasibility of such studies is questionable. Small, mostly retrospective, studies (Table 4) provide limited guidance on this issue; most studies focus on bleeding events rather than the cardiovascular efficacy of triple therapy. Because of these limitations, cardiovascular societies give IIb recommendation for either triple therapy or the combination of warfarin and clopidogrel in this setting and the level of evidence is C.1, 59, 60

In the largest study to date, Nguyen et al.58 evaluated 800 patients who underwent stenting for ACS and were discharged on warfarin plus single antiplatelet agent or triple therapy as part of the GRACE registry. At 6 months, triple therapy conferred a significant reduction in stroke (0.7% versus 3.4%, P = .02) but not in death or MI. There were no differences in in‐hospital major bleeding events between the two groups (5.9% versus 4.6%; P = .46). Similarly, Sarafoff et al.53 reported no significant differences in the combined endpoint (death, MI, stent thrombosis or stroke) or bleeding complications among patients who received triple therapy or DAPT at 2 years of follow‐up. In contrast, Ruiz‐Nodar et al.52 showed that triple therapy, compared with DAPT, at discharge reduced the incidence of death (17.8% versus 27.8%; adjusted HR = 3.43; 95% CI, 1.617.54; P = .002) and major adverse cardiac events (26.5% versus 38.7%; adjusted HR = 4.9; 95% CI, 2.1711.1; P = .01), without a substantial increase in major bleeding events.

The value of combination antiplatelet therapy to prevent stent thrombosis in these patients is clearer in the study reported by Karjalainen et al.10 This case‐control study of 239 patients receiving warfarin at baseline who underwent PCI evaluated a primary endpoint of death, MI, target‐vessel revascularization, or stent thrombosis and a secondary endpoint of major bleeding and stroke to 12 months of follow‐up. Forty‐eight percent of patients received triple therapy, whereas 15.5% were discharged on DAPT. The remaining patients received warfarin plus a single antiplatelet agent. Stent thrombosis occurred more frequently among patients receiving warfarin plus ASA (15.2%) than among those receiving triple therapy (1.9%). As expected, stroke was more frequent in patients treated with DAPT (8.8%) than among those receiving triple therapy (2.8%). Major bleeding was similar between groups. Therapy with warfarin was an independent predictor of both major bleeding and major cardiac events at 1 year. This observation illustrates that the outcome of PCI in patients on chronic warfarin therapy is unsatisfactory irrespective of the antithrombotic combinations used, highlighting the need for better strategies to treat these patients.

Choice of Therapy and Management of Patients Eligible for Triple Therapy

Current guidelines for PCI do not provide guidance for patients with an indication for triple therapy due to a paucity of published evidence. Several ongoing prospective trials aim to address the management of these patients (AFCAS, ISAR‐TRIPLE). Pending further study, clinicians should consider the embolic risk (CHADS2 score), target INR, type of stent, bleeding risk, and duration of treatment when determining the appropriate antiplatelet/anticoagulant combinations. The CHADS2 score (Table 2) stratifies the risk for stroke among patients with AF,4 while the Outpatient Bleeding Risk Index (OBRI) allows estimation of bleeding risk.12, 61 The OBRI considers age > 65 years, prior stroke, prior gastrointestinal bleeding, and any of four comorbidities (recent MI, anemia, diabetes, or renal insufficiency) in order to stratify patients into three risk groups.61 Patients with three to four risk factors have a high risk of bleeding (23% at 3 months and 48% at 12 months) whereas patients with no risk factors have only a 3% risk of bleeding at 12 months. Unfortunately, advanced age and prior stroke appear in both OBRI and CHADS.

For patients with AF who are at high risk for embolic stroke (>3% per year), we recommend triple therapy for the shortest time possible, followed by warfarin and ASA indefinitely. In case of BMS, it is acceptable to shorten triple therapy duration to 1 month. The optimal duration of triple therapy for patients with DES is uncertain; recommended durations range from 3 months to 1 year.62 If the potential consequences of stent thrombosis are high due to a large amount of myocardium at risk, an extended period of triple therapy might be justified. For patients whose stroke risk is lower (CHADS2 score of 0‐1), the risk for bleeding likely outweighs any benefit from stroke prevention. In this instance, it is reasonable to use DAPT with ASA and clopidogrel for 1 month after BMS and 12 months after DES, followed by ASA, with or without warfarin, indefinitely. In a recently published study, patients with AF and a CHADS2 score of 1 had a yearly stroke risk of 1.25% while taking DAPT63; the risk of major bleeding for triple therapy is 6.1% per year.64

For patients who have a high bleeding risk, BMS are the preferred stent type as the duration of triple therapy might be limited to 4 weeks. To our knowledge, no randomized study has evaluated the outcome of patients with BMS compared with DES who also have an indication for warfarin. Because studies have suggested that clopidogrel is more effective than aspirin in preventing stent thrombosis and in reducing death or MI after coronary stenting,40, 65 warfarin and single antiplatelet therapy with clopidogrel might be a reasonable treatment option in patients with high bleeding risk. The WOEST study (NCT00769938), currently recruiting participants, is the first randomized study specifically designed to test this hypothesis.

Since gastrointestinal bleeding accounts for approximately 30‐40% of hemorrhagic events in patients on combined ASA and anticoagulant therapy, an expert consensus document recommended concomitant treatment with proton pump inhibitors (PPIs) to reduce this risk.66 In contrast, the 2009 Focused Updates of the ACC/AHA/SCAI Guidelines did not recommend the use of PPIs with DAPT in the setting of ACS.2 This is because of studies that show inhibition of platelet activation,67 and potential clinical harm,68 when clopidogrel is combined with certain PPIs that inhibit the CYP2C19 enzyme. However, to date there are no convincing randomized clinical trial data documenting an important clinical drug‐drug interaction. The U.S. Food and Drug Administration (FDA) advises that physicians avoid the use of clopidogrel in patients with impaired CYP2C19 function due to known genetic variation or due to concomitant use of drugs that inhibit CYP2C19 activity. More specifically, the FDA recommends avoiding the use of omeprazole and esomeprazole in patients taking clopidogrel.69

In particular, elderly patients have an increased risk of bleeding while receiving triple therapy. In a study of patients over age 65, 2.5% were hospitalized for bleeding in the first year after PCI, and the use of triple therapy was the strongest predictor of bleeding (more than threefold increase).70 One in five patients suffered death or MI at 1 year after hospitalization for bleeding.70 The basis for poor outcomes after hospitalization for bleeding in this population is multifactorial and may be due to the location of bleeding, associated hypercoagulable state, potential adverse impact of blood transfusion, withdrawal of warfarin therapy in patients with AF and PCI, and the premature discontinuation of DAPT. The use of nonsteroidal anti‐inflammatory drugs (NSAIDs) is common among the elderly and conferred a doubling of bleeding risk.70 Limiting the use of NSAID, the use of low‐dose ASA beyond 30 days after stent implantation, greater use of BMS, and maintaining INR at the lowest possible level (INR of 22.5) will reduce the risk for bleeding.57, 71

New Anticoagulants

Due to the high risk for bleeding with warfarin and the challenges inherent in INR monitoring, researchers have developed several novel anticoagulants whose advantages include fixed daily dosing and no need for monitoring. Dabigatran is a direct oral thrombin inhibitor that is already licensed in Europe and Canada for thromboprophylaxis after hip or knee surgery. It has also been studied in patients with AF. In the RE‐LY trial, patients with AF who received dabigatran 110 mg daily had rates of stroke and systemic embolism that were similar to those with warfarin, as well as lower rates of major hemorrhage.72 The randomized ReDEEM trial, reported at the AHA 2009 Scientific Sessions, was aimed at finding a dosage of dabigatran that achieves a good balance between clinical effectiveness and bleeding risk when combined with aspirin and clopidogrel after acute MI. Dosages ranging from 50 mg twice daily to 150 mg twice daily were all associated with 6‐month rates of bleeding lower than 2%. Hospitalists should view these encouraging results cautiously until the publication of ReDEEM trial results in a peer‐reviewed journal.

A variety of oral Xa antagonists are also being evaluated in patients with AF or ACS. These trials offer insight into triple therapy regimens that include ASA, clopidogrel, and an Xa antagonist. In a recent study of the oral Xa antagonist rivaroxaban, investigators stratified 3491 subjects with ACS according to whether they received concomitant ASA alone or ASA and clopidogrel.73 Subjects receiving ASA plus rivaroxaban had a modest increase in bleeding. Triple therapy, however, increased the composite bleeding rate from 3.5% in the DAPT group to approximately 6‐15% (low‐dose or high‐dose rivaroxaban, respectively). Rivaroxaban is currently under review by the FDA.

These novel agents might eventually replace warfarin for many or most indications for anticoagulation. It is imperative that future research compare the efficacy and risk of bleeding between triple therapy using these new agents and triple therapy with warfarin.

Conclusions

The management of patients on long‐term anticoagulation who require DAPT because of ACS or coronary stenting is challenging. DAPT may safely substitute for warfarin only for patients at low risk for a thromboembolic event (ie, low‐risk AF with low CHADS2 score). Clinicians should not interrupt warfarin in patients at higher risk (ie, intermediate to high‐risk AF, mechanical valves, or recent venous thromboembolism), even in the presence of DAPT. In these patients, triple therapy is the optimal approach following coronary stenting (and possibly during the initial period after ACS without stenting). As this approach confers a fivefold increase in bleeding complications compared with DAPT, careful monitoring of the INR, the addition of PPIs, and the exclusion of elderly patients who are at the highest risk for bleeding complications74 is recommended. The preferred duration of triple therapy after BMS in patients who require long‐term anticoagulation is 1 month, whereas the optimal duration after ACS or DES remains unresolved.