Acute pulmonary embolism (PE) and acute myocardial infarction (AMI) are common inpatient diagnoses, and are frequently in the differential diagnosis of patients evaluated for chest pain and dyspnea. We present a case with 1 unifying explanation for these entities to coexist. Acute PE with subsequent embolism to the coronary arteries via a patent foramen ovale (PFO) is rare, but the underlying disorder and anatomical variant are common. Of practical significance, hospitalized patients with acute PE and PFO may have up to a 5‐fold increase in morbidity compared to patients with isolated PE.

CASE REPORT

A 79‐year‐old male smoker underwent resection of a recurrent high‐grade liposarcoma of the right upper extremity. He had no antecedent history of coronary artery disease (CAD) or atrial fibrillation, and had no additional vascular risk factors. On postoperative day 2, he developed acute chest pain, dyspnea, and hypoxia. He appeared alert but was diaphoretic and in moderate distress. Pulse was 84 beats per minute, blood pressure 230/120 mm Hg, and oxygen saturation 59% on room air (93% on supplemental oxygen). Heart and lung exam were unremarkable. Neck veins were not distended. Extremity exam was negative for edema, asymmetry, or calf tenderness, and pedal pulses were palpable bilaterally.

The patient's initial complete blood count, metabolic panel, and cardiac enzymes were within normal limits. Arterial blood gas (on 4‐L nasal cannula) revealed pH of 7.35, partial pressure of arterial oxygen (PaO₂) of 65.5 mm Hg, partial pressure of arterial CO₂ (PaCO₂) of 45.4 mm Hg, and an alveolar‐arterial gradient of 131.3 mm Hg. Electrocardiogram (ECG) (Figure 1) showed an unchanged right bundle branch block, but new 2.5‐mm ST segment elevation in leads V4‐6, III, and aVF, and ST depression in aVL. At this point, the available data suggested either PE with secondary ECG changes or acute ST‐elevation MI with hypoxia. Given the ST elevation in 2 coronary distributions and concern for multivessel CAD, the patient was referred for emergent coronary angiography.

Figure 1

The patient was given aspirin, intravenous unfractionated heparin, and morphine. Left heart catheterization showed an abrupt cutoff in the distal left anterior descending artery (LAD), suggestive of thrombosis secondary to coronary embolism (Figure 2A); angioplasty was not attempted due to the distal location of the occlusion. The posterior descending artery from the right coronary artery was relatively short and the inferior apex was supplied by the distal LAD. The left ventriculogram demonstrated preserved ejection fraction but severe apical hypokinesis, correlating with the occluded vascular territory. The remaining coronary arteries were without significant stenosis. Based on the angiogram findings, paradoxical embolism was suspected. Right heart catheterization identified a previously undiagnosed PFO (Figure 2B); no thrombus was visualized. Right to left shunt was not identified; right ventricular systolic pressure was 46 mm Hg (normal, 15‐25 mm Hg). Subsequent spiral computed tomography (CT) revealed bilateral PEs. It was concluded that the patient had suffered an acute PE, from which the thrombus was able to traverse the PFO and left heart, ultimately entering the LAD, causing the acute embolic MI (Figure 3). ST elevation was present in the inferior leads secondary to the wraparound LAD that supplied the inferior apex, as demonstrated by the wall motion abnormality present on ventriculography.

Figure 2

Figure 3

The patient was felt to be in a hypercoagulable state due to his malignancy and recent surgery; given the diagnosis of an acute PE, no ultrasound was performed to search for a deep vein thrombosis. The patient was transferred to the intensive care unit, continued on intravenous heparin and oxygen and started on oral warfarin. Subsequent hospital course was complicated by aspiration pneumonia and worsening hypoxia, but after a 17‐day hospital stay he was weaned off supplemental oxygen and transferred to an extended care facility with a therapeutic international normalized ratio (INR).

DISCUSSION

PFOs are congenital cardiac lesions that may persist through adulthood1 and are found incidentally in 19% to 36% of the normal population.2 They may range from 1 to 19 mm in size.2, 3 Contrast echocardiography has enabled a simple, accurate, and safe procedure for diagnosis (>5 microbubbles in the left heart cavities within three cardiac cycles after their appearance in the right atrium is considered diagnostic), though transesophageal echocardiography (TEE) is the gold standard.4

First described by Cohnheim in 1877,5 paradoxical embolism refers to the passage of embolic material from the venous to arterial circulation through a cardiac defect such as a PFO. However, definite confirmation of paradoxical embolism essentially requires catching the thrombus in the act of crossing the foramen ovale. Direct observation of this during life is rarely possible, and remains confined to isolated echocardiographic reports.6‐13

In clinical practice, the diagnosis of paradoxical embolism is almost always presumptive and relies on: (1) the occurrence of an arterial thromboembolic event in the absence of atrial fibrillation, left‐sided heart disease, or severe atherosclerosis; (2) the detection of right‐to‐left shunt, usually through a PFO or an atrial septal defect (ASD); and (3) the presence of venous thrombosis or PE.14

Although most patients are asymptomatic, during the past 20 years an association of PFO with stroke, migraine headache, peripheral arterial occlusion, and decompression induced neurologic dysfunction has been suggested.1 The neurological symptoms are proposed to be secondary to passage of small thrombi from the venous system through the PFO into arterial circulation during a transient right‐to‐left shunt. The source of clots cannot be established in most patients; fewer than 10% will have detectable deep vein thrombosis.15 Even less common are paradoxical emboli to the coronary arteries,1 estimated at 5% to 10% of all paradoxical emboli.16

In order for a thrombus to paradoxically embolize across a PFO, an atrial right‐to‐left pressure gradient must be present. Such a gradient occurs in normal individuals during early ventricular systole and with a Valsalva maneuver.17‐19 In a community‐based cohort study conducted to evaluate potential stroke risk, 148 (out of 581) subjects were found to have a PFO by TEE; 84 (57%) had right‐to‐left shunting at rest, and 136 (92%) had right‐to‐left shunting with Valsalva.2 Pathologic instances of pulmonary hypertension such as PE further elevate right‐heart pressures, further promoting intracardiac shunt.

Acute PE in the setting of PFO carries important prognostic implications. Konstantinides et al.20 prospectively evaluated 139 patients with large acute PE, all of whom had pulmonary hypertension and 96% of whom had right ventricular dilatation. They found a high prevalence of PFOs (35%) in this population. Furthermore, the subgroup with both PE and PFO had a high mortality (33% death rate compared with 14% in those without PFO; P = 0.015). When logistic regression analysis was performed, only arterial hypotension (odds ratio [OR] 26.3; P < 0.001) and the presence of PFO (OR 11.4; P < 0.001) remained significantly correlated with mortality. The authors reported that the presence of a PFO was associated with more than a 5‐fold increase in the adjusted risk of major in‐hospital complications (P < 0.001); no specific etiologic factors were proposed for this association.

In general, MI in the absence of CAD is uncommon, comprising <1% to 6% of all cases.21 No cause is found for the majority, but reported etiologies include coronary spasm in 15%, hypercoagulable states in 13%, collagen vascular diseases in 2%, and paradoxical embolism in 2%.22 AMI due to coronary embolism is uncommon, and when it does occur, left‐to‐left emboli in the setting of atrial fibrillation or prosthetic valves are far more common than paradoxical emboli. In an autopsy series of 1,050 patients with MIs, Prizel et al.23 identified only 55 patients with coronary embolism, none of which was right‐sided in origin.

A handful of published case reports documented paradoxical embolism as a cause for AMI.24‐26 Reported cases more often involved an ASD rather than PFO.27, 28 In most cases, diagnosis was made postmortem, though in a comprehensive review of the literature, Meier‐Ewert et al.13 identified 8 cases of AMI due to paradoxical embolism being diagnosed antemortem. Paradoxical emboli have been identified in all major divisions of the epicardial circulation, though involvement of the left coronary circulation is more common than the right.16

It is well‐established that the prevalence of PFO in patients with cryptogenic stroke is significantly higher than in the general population,1 and Crump et al.21 examined a case‐matched series of 18 patients with AMI who had little to no CAD (<30% stenosis) to see if the frequency of PFO was similarly higher in this group. Each group had identical frequency of PFO (28%; P = NS). The authors concluded that PFO is unlikely to contribute significantly to AMI. However, this study was limited by the small number of patients and the fact that transthoracic echocardiography (TTE) was used instead of TEE for the diagnosis of PFO.

The definitive diagnosis of AMI due to paradoxical embolism requires angiographic findings consistent with embolic occlusion (such as the cutoff sign in a distal coronary artery we observed in Figure 2A), cardiac defects predisposing to paradoxical emboli (such as PFO), and evidence of a venous source for thromboembolism. Alternatively, diagnosis can be made via direct visualization of emboli in the coronary arteries by TEE, or by autopsy.

Although electrocardiography is essential in the diagnosis and treatment of MI, it has the potential to be deceptive. Acute pulmonary hypertension caused by PE may be accompanied by ST elevation in inferior leads II, III, and aV_F in a pseudoinfarction pattern mimicking AMI.29 This ECG abnormality probably reflects reciprocal changes of inferoposterior ischemia from right ventricular pressure overloading. However, clearly distinguishing between pseudoinfarction and true inferior infarction in the setting of PE requires coronary angiography.

Regarding therapy, acute treatment of PE is well‐established and consists of at least 5 days of therapeutically‐dosed heparin product that overlaps with therapeutic warfarin anticoagulation. Management of the PFO and coronary embolism is less clear; there are no guidelines for treatment of coronary embolism. Management strategies should focus on treatment of acute ischemia as well as prevention of future emboli, principally anticoagulation. Because the pathogenesis of AMI in this setting is drastically different from MI secondary to atherosclerosis, there is neither a biological basis nor clinical data to suggest benefit from initiation of beta‐blockers, aspirin, angiotensin converting enzyme inhibitors, or statins.

While studies have been done, and are underway to address optimal management of PFO in the setting of both stroke and migraine headache, to our knowledge, no such trials have addressed PFO and MI. Mehan et al.30 reported 2 cases of AMI caused by suspected paradoxical embolism, and in both cases, instant percutaneous closure of PFO was undertaken. However, there are no data to support or refute such an intervention in this particular setting.

Acute pulmonary embolism (PE) and acute myocardial infarction (AMI) are common inpatient diagnoses, and are frequently in the differential diagnosis of patients evaluated for chest pain and dyspnea. We present a case with 1 unifying explanation for these entities to coexist. Acute PE with subsequent embolism to the coronary arteries via a patent foramen ovale (PFO) is rare, but the underlying disorder and anatomical variant are common. Of practical significance, hospitalized patients with acute PE and PFO may have up to a 5‐fold increase in morbidity compared to patients with isolated PE.

CASE REPORT

A 79‐year‐old male smoker underwent resection of a recurrent high‐grade liposarcoma of the right upper extremity. He had no antecedent history of coronary artery disease (CAD) or atrial fibrillation, and had no additional vascular risk factors. On postoperative day 2, he developed acute chest pain, dyspnea, and hypoxia. He appeared alert but was diaphoretic and in moderate distress. Pulse was 84 beats per minute, blood pressure 230/120 mm Hg, and oxygen saturation 59% on room air (93% on supplemental oxygen). Heart and lung exam were unremarkable. Neck veins were not distended. Extremity exam was negative for edema, asymmetry, or calf tenderness, and pedal pulses were palpable bilaterally.

The patient's initial complete blood count, metabolic panel, and cardiac enzymes were within normal limits. Arterial blood gas (on 4‐L nasal cannula) revealed pH of 7.35, partial pressure of arterial oxygen (PaO₂) of 65.5 mm Hg, partial pressure of arterial CO₂ (PaCO₂) of 45.4 mm Hg, and an alveolar‐arterial gradient of 131.3 mm Hg. Electrocardiogram (ECG) (Figure 1) showed an unchanged right bundle branch block, but new 2.5‐mm ST segment elevation in leads V4‐6, III, and aVF, and ST depression in aVL. At this point, the available data suggested either PE with secondary ECG changes or acute ST‐elevation MI with hypoxia. Given the ST elevation in 2 coronary distributions and concern for multivessel CAD, the patient was referred for emergent coronary angiography.

Figure 1

The patient was given aspirin, intravenous unfractionated heparin, and morphine. Left heart catheterization showed an abrupt cutoff in the distal left anterior descending artery (LAD), suggestive of thrombosis secondary to coronary embolism (Figure 2A); angioplasty was not attempted due to the distal location of the occlusion. The posterior descending artery from the right coronary artery was relatively short and the inferior apex was supplied by the distal LAD. The left ventriculogram demonstrated preserved ejection fraction but severe apical hypokinesis, correlating with the occluded vascular territory. The remaining coronary arteries were without significant stenosis. Based on the angiogram findings, paradoxical embolism was suspected. Right heart catheterization identified a previously undiagnosed PFO (Figure 2B); no thrombus was visualized. Right to left shunt was not identified; right ventricular systolic pressure was 46 mm Hg (normal, 15‐25 mm Hg). Subsequent spiral computed tomography (CT) revealed bilateral PEs. It was concluded that the patient had suffered an acute PE, from which the thrombus was able to traverse the PFO and left heart, ultimately entering the LAD, causing the acute embolic MI (Figure 3). ST elevation was present in the inferior leads secondary to the wraparound LAD that supplied the inferior apex, as demonstrated by the wall motion abnormality present on ventriculography.

Figure 2

Figure 3

The patient was felt to be in a hypercoagulable state due to his malignancy and recent surgery; given the diagnosis of an acute PE, no ultrasound was performed to search for a deep vein thrombosis. The patient was transferred to the intensive care unit, continued on intravenous heparin and oxygen and started on oral warfarin. Subsequent hospital course was complicated by aspiration pneumonia and worsening hypoxia, but after a 17‐day hospital stay he was weaned off supplemental oxygen and transferred to an extended care facility with a therapeutic international normalized ratio (INR).

DISCUSSION

PFOs are congenital cardiac lesions that may persist through adulthood1 and are found incidentally in 19% to 36% of the normal population.2 They may range from 1 to 19 mm in size.2, 3 Contrast echocardiography has enabled a simple, accurate, and safe procedure for diagnosis (>5 microbubbles in the left heart cavities within three cardiac cycles after their appearance in the right atrium is considered diagnostic), though transesophageal echocardiography (TEE) is the gold standard.4

First described by Cohnheim in 1877,5 paradoxical embolism refers to the passage of embolic material from the venous to arterial circulation through a cardiac defect such as a PFO. However, definite confirmation of paradoxical embolism essentially requires catching the thrombus in the act of crossing the foramen ovale. Direct observation of this during life is rarely possible, and remains confined to isolated echocardiographic reports.6‐13

In clinical practice, the diagnosis of paradoxical embolism is almost always presumptive and relies on: (1) the occurrence of an arterial thromboembolic event in the absence of atrial fibrillation, left‐sided heart disease, or severe atherosclerosis; (2) the detection of right‐to‐left shunt, usually through a PFO or an atrial septal defect (ASD); and (3) the presence of venous thrombosis or PE.14

Although most patients are asymptomatic, during the past 20 years an association of PFO with stroke, migraine headache, peripheral arterial occlusion, and decompression induced neurologic dysfunction has been suggested.1 The neurological symptoms are proposed to be secondary to passage of small thrombi from the venous system through the PFO into arterial circulation during a transient right‐to‐left shunt. The source of clots cannot be established in most patients; fewer than 10% will have detectable deep vein thrombosis.15 Even less common are paradoxical emboli to the coronary arteries,1 estimated at 5% to 10% of all paradoxical emboli.16

In order for a thrombus to paradoxically embolize across a PFO, an atrial right‐to‐left pressure gradient must be present. Such a gradient occurs in normal individuals during early ventricular systole and with a Valsalva maneuver.17‐19 In a community‐based cohort study conducted to evaluate potential stroke risk, 148 (out of 581) subjects were found to have a PFO by TEE; 84 (57%) had right‐to‐left shunting at rest, and 136 (92%) had right‐to‐left shunting with Valsalva.2 Pathologic instances of pulmonary hypertension such as PE further elevate right‐heart pressures, further promoting intracardiac shunt.

Acute PE in the setting of PFO carries important prognostic implications. Konstantinides et al.20 prospectively evaluated 139 patients with large acute PE, all of whom had pulmonary hypertension and 96% of whom had right ventricular dilatation. They found a high prevalence of PFOs (35%) in this population. Furthermore, the subgroup with both PE and PFO had a high mortality (33% death rate compared with 14% in those without PFO; P = 0.015). When logistic regression analysis was performed, only arterial hypotension (odds ratio [OR] 26.3; P < 0.001) and the presence of PFO (OR 11.4; P < 0.001) remained significantly correlated with mortality. The authors reported that the presence of a PFO was associated with more than a 5‐fold increase in the adjusted risk of major in‐hospital complications (P < 0.001); no specific etiologic factors were proposed for this association.

In general, MI in the absence of CAD is uncommon, comprising <1% to 6% of all cases.21 No cause is found for the majority, but reported etiologies include coronary spasm in 15%, hypercoagulable states in 13%, collagen vascular diseases in 2%, and paradoxical embolism in 2%.22 AMI due to coronary embolism is uncommon, and when it does occur, left‐to‐left emboli in the setting of atrial fibrillation or prosthetic valves are far more common than paradoxical emboli. In an autopsy series of 1,050 patients with MIs, Prizel et al.23 identified only 55 patients with coronary embolism, none of which was right‐sided in origin.

A handful of published case reports documented paradoxical embolism as a cause for AMI.24‐26 Reported cases more often involved an ASD rather than PFO.27, 28 In most cases, diagnosis was made postmortem, though in a comprehensive review of the literature, Meier‐Ewert et al.13 identified 8 cases of AMI due to paradoxical embolism being diagnosed antemortem. Paradoxical emboli have been identified in all major divisions of the epicardial circulation, though involvement of the left coronary circulation is more common than the right.16

It is well‐established that the prevalence of PFO in patients with cryptogenic stroke is significantly higher than in the general population,1 and Crump et al.21 examined a case‐matched series of 18 patients with AMI who had little to no CAD (<30% stenosis) to see if the frequency of PFO was similarly higher in this group. Each group had identical frequency of PFO (28%; P = NS). The authors concluded that PFO is unlikely to contribute significantly to AMI. However, this study was limited by the small number of patients and the fact that transthoracic echocardiography (TTE) was used instead of TEE for the diagnosis of PFO.

The definitive diagnosis of AMI due to paradoxical embolism requires angiographic findings consistent with embolic occlusion (such as the cutoff sign in a distal coronary artery we observed in Figure 2A), cardiac defects predisposing to paradoxical emboli (such as PFO), and evidence of a venous source for thromboembolism. Alternatively, diagnosis can be made via direct visualization of emboli in the coronary arteries by TEE, or by autopsy.

Although electrocardiography is essential in the diagnosis and treatment of MI, it has the potential to be deceptive. Acute pulmonary hypertension caused by PE may be accompanied by ST elevation in inferior leads II, III, and aV_F in a pseudoinfarction pattern mimicking AMI.29 This ECG abnormality probably reflects reciprocal changes of inferoposterior ischemia from right ventricular pressure overloading. However, clearly distinguishing between pseudoinfarction and true inferior infarction in the setting of PE requires coronary angiography.

Regarding therapy, acute treatment of PE is well‐established and consists of at least 5 days of therapeutically‐dosed heparin product that overlaps with therapeutic warfarin anticoagulation. Management of the PFO and coronary embolism is less clear; there are no guidelines for treatment of coronary embolism. Management strategies should focus on treatment of acute ischemia as well as prevention of future emboli, principally anticoagulation. Because the pathogenesis of AMI in this setting is drastically different from MI secondary to atherosclerosis, there is neither a biological basis nor clinical data to suggest benefit from initiation of beta‐blockers, aspirin, angiotensin converting enzyme inhibitors, or statins.

While studies have been done, and are underway to address optimal management of PFO in the setting of both stroke and migraine headache, to our knowledge, no such trials have addressed PFO and MI. Mehan et al.30 reported 2 cases of AMI caused by suspected paradoxical embolism, and in both cases, instant percutaneous closure of PFO was undertaken. However, there are no data to support or refute such an intervention in this particular setting.

Acute pulmonary embolism (PE) and acute myocardial infarction (AMI) are common inpatient diagnoses, and are frequently in the differential diagnosis of patients evaluated for chest pain and dyspnea. We present a case with 1 unifying explanation for these entities to coexist. Acute PE with subsequent embolism to the coronary arteries via a patent foramen ovale (PFO) is rare, but the underlying disorder and anatomical variant are common. Of practical significance, hospitalized patients with acute PE and PFO may have up to a 5‐fold increase in morbidity compared to patients with isolated PE.

CASE REPORT

A 79‐year‐old male smoker underwent resection of a recurrent high‐grade liposarcoma of the right upper extremity. He had no antecedent history of coronary artery disease (CAD) or atrial fibrillation, and had no additional vascular risk factors. On postoperative day 2, he developed acute chest pain, dyspnea, and hypoxia. He appeared alert but was diaphoretic and in moderate distress. Pulse was 84 beats per minute, blood pressure 230/120 mm Hg, and oxygen saturation 59% on room air (93% on supplemental oxygen). Heart and lung exam were unremarkable. Neck veins were not distended. Extremity exam was negative for edema, asymmetry, or calf tenderness, and pedal pulses were palpable bilaterally.

The patient's initial complete blood count, metabolic panel, and cardiac enzymes were within normal limits. Arterial blood gas (on 4‐L nasal cannula) revealed pH of 7.35, partial pressure of arterial oxygen (PaO₂) of 65.5 mm Hg, partial pressure of arterial CO₂ (PaCO₂) of 45.4 mm Hg, and an alveolar‐arterial gradient of 131.3 mm Hg. Electrocardiogram (ECG) (Figure 1) showed an unchanged right bundle branch block, but new 2.5‐mm ST segment elevation in leads V4‐6, III, and aVF, and ST depression in aVL. At this point, the available data suggested either PE with secondary ECG changes or acute ST‐elevation MI with hypoxia. Given the ST elevation in 2 coronary distributions and concern for multivessel CAD, the patient was referred for emergent coronary angiography.

Figure 1

The patient was given aspirin, intravenous unfractionated heparin, and morphine. Left heart catheterization showed an abrupt cutoff in the distal left anterior descending artery (LAD), suggestive of thrombosis secondary to coronary embolism (Figure 2A); angioplasty was not attempted due to the distal location of the occlusion. The posterior descending artery from the right coronary artery was relatively short and the inferior apex was supplied by the distal LAD. The left ventriculogram demonstrated preserved ejection fraction but severe apical hypokinesis, correlating with the occluded vascular territory. The remaining coronary arteries were without significant stenosis. Based on the angiogram findings, paradoxical embolism was suspected. Right heart catheterization identified a previously undiagnosed PFO (Figure 2B); no thrombus was visualized. Right to left shunt was not identified; right ventricular systolic pressure was 46 mm Hg (normal, 15‐25 mm Hg). Subsequent spiral computed tomography (CT) revealed bilateral PEs. It was concluded that the patient had suffered an acute PE, from which the thrombus was able to traverse the PFO and left heart, ultimately entering the LAD, causing the acute embolic MI (Figure 3). ST elevation was present in the inferior leads secondary to the wraparound LAD that supplied the inferior apex, as demonstrated by the wall motion abnormality present on ventriculography.

Figure 2

Figure 3

The patient was felt to be in a hypercoagulable state due to his malignancy and recent surgery; given the diagnosis of an acute PE, no ultrasound was performed to search for a deep vein thrombosis. The patient was transferred to the intensive care unit, continued on intravenous heparin and oxygen and started on oral warfarin. Subsequent hospital course was complicated by aspiration pneumonia and worsening hypoxia, but after a 17‐day hospital stay he was weaned off supplemental oxygen and transferred to an extended care facility with a therapeutic international normalized ratio (INR).

DISCUSSION

PFOs are congenital cardiac lesions that may persist through adulthood1 and are found incidentally in 19% to 36% of the normal population.2 They may range from 1 to 19 mm in size.2, 3 Contrast echocardiography has enabled a simple, accurate, and safe procedure for diagnosis (>5 microbubbles in the left heart cavities within three cardiac cycles after their appearance in the right atrium is considered diagnostic), though transesophageal echocardiography (TEE) is the gold standard.4

First described by Cohnheim in 1877,5 paradoxical embolism refers to the passage of embolic material from the venous to arterial circulation through a cardiac defect such as a PFO. However, definite confirmation of paradoxical embolism essentially requires catching the thrombus in the act of crossing the foramen ovale. Direct observation of this during life is rarely possible, and remains confined to isolated echocardiographic reports.6‐13

In clinical practice, the diagnosis of paradoxical embolism is almost always presumptive and relies on: (1) the occurrence of an arterial thromboembolic event in the absence of atrial fibrillation, left‐sided heart disease, or severe atherosclerosis; (2) the detection of right‐to‐left shunt, usually through a PFO or an atrial septal defect (ASD); and (3) the presence of venous thrombosis or PE.14

Although most patients are asymptomatic, during the past 20 years an association of PFO with stroke, migraine headache, peripheral arterial occlusion, and decompression induced neurologic dysfunction has been suggested.1 The neurological symptoms are proposed to be secondary to passage of small thrombi from the venous system through the PFO into arterial circulation during a transient right‐to‐left shunt. The source of clots cannot be established in most patients; fewer than 10% will have detectable deep vein thrombosis.15 Even less common are paradoxical emboli to the coronary arteries,1 estimated at 5% to 10% of all paradoxical emboli.16

In order for a thrombus to paradoxically embolize across a PFO, an atrial right‐to‐left pressure gradient must be present. Such a gradient occurs in normal individuals during early ventricular systole and with a Valsalva maneuver.17‐19 In a community‐based cohort study conducted to evaluate potential stroke risk, 148 (out of 581) subjects were found to have a PFO by TEE; 84 (57%) had right‐to‐left shunting at rest, and 136 (92%) had right‐to‐left shunting with Valsalva.2 Pathologic instances of pulmonary hypertension such as PE further elevate right‐heart pressures, further promoting intracardiac shunt.

Acute PE in the setting of PFO carries important prognostic implications. Konstantinides et al.20 prospectively evaluated 139 patients with large acute PE, all of whom had pulmonary hypertension and 96% of whom had right ventricular dilatation. They found a high prevalence of PFOs (35%) in this population. Furthermore, the subgroup with both PE and PFO had a high mortality (33% death rate compared with 14% in those without PFO; P = 0.015). When logistic regression analysis was performed, only arterial hypotension (odds ratio [OR] 26.3; P < 0.001) and the presence of PFO (OR 11.4; P < 0.001) remained significantly correlated with mortality. The authors reported that the presence of a PFO was associated with more than a 5‐fold increase in the adjusted risk of major in‐hospital complications (P < 0.001); no specific etiologic factors were proposed for this association.

In general, MI in the absence of CAD is uncommon, comprising <1% to 6% of all cases.21 No cause is found for the majority, but reported etiologies include coronary spasm in 15%, hypercoagulable states in 13%, collagen vascular diseases in 2%, and paradoxical embolism in 2%.22 AMI due to coronary embolism is uncommon, and when it does occur, left‐to‐left emboli in the setting of atrial fibrillation or prosthetic valves are far more common than paradoxical emboli. In an autopsy series of 1,050 patients with MIs, Prizel et al.23 identified only 55 patients with coronary embolism, none of which was right‐sided in origin.

A handful of published case reports documented paradoxical embolism as a cause for AMI.24‐26 Reported cases more often involved an ASD rather than PFO.27, 28 In most cases, diagnosis was made postmortem, though in a comprehensive review of the literature, Meier‐Ewert et al.13 identified 8 cases of AMI due to paradoxical embolism being diagnosed antemortem. Paradoxical emboli have been identified in all major divisions of the epicardial circulation, though involvement of the left coronary circulation is more common than the right.16

It is well‐established that the prevalence of PFO in patients with cryptogenic stroke is significantly higher than in the general population,1 and Crump et al.21 examined a case‐matched series of 18 patients with AMI who had little to no CAD (<30% stenosis) to see if the frequency of PFO was similarly higher in this group. Each group had identical frequency of PFO (28%; P = NS). The authors concluded that PFO is unlikely to contribute significantly to AMI. However, this study was limited by the small number of patients and the fact that transthoracic echocardiography (TTE) was used instead of TEE for the diagnosis of PFO.

The definitive diagnosis of AMI due to paradoxical embolism requires angiographic findings consistent with embolic occlusion (such as the cutoff sign in a distal coronary artery we observed in Figure 2A), cardiac defects predisposing to paradoxical emboli (such as PFO), and evidence of a venous source for thromboembolism. Alternatively, diagnosis can be made via direct visualization of emboli in the coronary arteries by TEE, or by autopsy.

Although electrocardiography is essential in the diagnosis and treatment of MI, it has the potential to be deceptive. Acute pulmonary hypertension caused by PE may be accompanied by ST elevation in inferior leads II, III, and aV_F in a pseudoinfarction pattern mimicking AMI.29 This ECG abnormality probably reflects reciprocal changes of inferoposterior ischemia from right ventricular pressure overloading. However, clearly distinguishing between pseudoinfarction and true inferior infarction in the setting of PE requires coronary angiography.

Regarding therapy, acute treatment of PE is well‐established and consists of at least 5 days of therapeutically‐dosed heparin product that overlaps with therapeutic warfarin anticoagulation. Management of the PFO and coronary embolism is less clear; there are no guidelines for treatment of coronary embolism. Management strategies should focus on treatment of acute ischemia as well as prevention of future emboli, principally anticoagulation. Because the pathogenesis of AMI in this setting is drastically different from MI secondary to atherosclerosis, there is neither a biological basis nor clinical data to suggest benefit from initiation of beta‐blockers, aspirin, angiotensin converting enzyme inhibitors, or statins.

While studies have been done, and are underway to address optimal management of PFO in the setting of both stroke and migraine headache, to our knowledge, no such trials have addressed PFO and MI. Mehan et al.30 reported 2 cases of AMI caused by suspected paradoxical embolism, and in both cases, instant percutaneous closure of PFO was undertaken. However, there are no data to support or refute such an intervention in this particular setting.

A previously healthy 50‐year‐old man was eating a meal of rigatoni and shrimp at his favorite San Francisco restaurant when he suddenly developed severe pain in his throat followed a short time later by pleuritic chest pain localized to the anterior right chest. He completed his meal and then sought medical attention in the Emergency Department. He was in mild distress secondary to pain but his physical examination was otherwise unremarkable. His laboratory studies showed a white blood cell count of 15,300/mm³ with 86% neutrophils. A chest X‐ray, electrocardiogram, cardiac enzymes, and ventilation/perfusion scan were all normal. Because there was suspicion of an ingested foreign body, an abdominal radiograph was obtained which revealed a 2‐cm trapezoidal foreign body in the right lower quadrant (Figure 1; arrow). A chest computed tomography (CT) scan revealed air in the mediastinum consistent with esophageal perforation (Figure 2). One day after admission a Hypaque esophagram showed trauma to the posterior cervical region of the esophagus, but no leak of contrast material into the mediastinum. The patient was managed conservatively with intravenous (IV) antibiotics and nothing by mouth. Stools were screened and 48 hours after admission the patient (painlessly) passed a piece of glass with a very sharp point (Figure 3) correlating in size and shape to the foreign body seen on the previous abdominal radiograph. The glass had apparently fallen into the patient's restaurant meal the night of admission. The patient did well and was discharged 6 days after admission. He asked the manager of his favorite restaurant to reimburse him the $200 copayment required for hospitalization required under his Preferred Provider Plan; the request was immediately honored.

Figure 1

Figure 2

Figure 3

Esophageal perforation is an emergency because of its high mortality rate. The most frequent cause is iatrogenic with instrumentation from endoscopic procedures. Other causes include foreign body ingestion (as in this case), trauma, operative injury, and tumor.1 Aggressive surgical intervention vs. conservative nonsurgical management remains a controversial topic.2 Early‐contained perforations can be managed successfully by limiting oral intake and giving parenteral antibiotics.3, 4 Any signs of sepsis, deterioration in the patient's condition, or uncontained rupture warrants immediate surgical intervention.14

A previously healthy 50‐year‐old man was eating a meal of rigatoni and shrimp at his favorite San Francisco restaurant when he suddenly developed severe pain in his throat followed a short time later by pleuritic chest pain localized to the anterior right chest. He completed his meal and then sought medical attention in the Emergency Department. He was in mild distress secondary to pain but his physical examination was otherwise unremarkable. His laboratory studies showed a white blood cell count of 15,300/mm³ with 86% neutrophils. A chest X‐ray, electrocardiogram, cardiac enzymes, and ventilation/perfusion scan were all normal. Because there was suspicion of an ingested foreign body, an abdominal radiograph was obtained which revealed a 2‐cm trapezoidal foreign body in the right lower quadrant (Figure 1; arrow). A chest computed tomography (CT) scan revealed air in the mediastinum consistent with esophageal perforation (Figure 2). One day after admission a Hypaque esophagram showed trauma to the posterior cervical region of the esophagus, but no leak of contrast material into the mediastinum. The patient was managed conservatively with intravenous (IV) antibiotics and nothing by mouth. Stools were screened and 48 hours after admission the patient (painlessly) passed a piece of glass with a very sharp point (Figure 3) correlating in size and shape to the foreign body seen on the previous abdominal radiograph. The glass had apparently fallen into the patient's restaurant meal the night of admission. The patient did well and was discharged 6 days after admission. He asked the manager of his favorite restaurant to reimburse him the $200 copayment required for hospitalization required under his Preferred Provider Plan; the request was immediately honored.

Figure 1

Figure 2

Figure 3

Esophageal perforation is an emergency because of its high mortality rate. The most frequent cause is iatrogenic with instrumentation from endoscopic procedures. Other causes include foreign body ingestion (as in this case), trauma, operative injury, and tumor.1 Aggressive surgical intervention vs. conservative nonsurgical management remains a controversial topic.2 Early‐contained perforations can be managed successfully by limiting oral intake and giving parenteral antibiotics.3, 4 Any signs of sepsis, deterioration in the patient's condition, or uncontained rupture warrants immediate surgical intervention.14

A previously healthy 50‐year‐old man was eating a meal of rigatoni and shrimp at his favorite San Francisco restaurant when he suddenly developed severe pain in his throat followed a short time later by pleuritic chest pain localized to the anterior right chest. He completed his meal and then sought medical attention in the Emergency Department. He was in mild distress secondary to pain but his physical examination was otherwise unremarkable. His laboratory studies showed a white blood cell count of 15,300/mm³ with 86% neutrophils. A chest X‐ray, electrocardiogram, cardiac enzymes, and ventilation/perfusion scan were all normal. Because there was suspicion of an ingested foreign body, an abdominal radiograph was obtained which revealed a 2‐cm trapezoidal foreign body in the right lower quadrant (Figure 1; arrow). A chest computed tomography (CT) scan revealed air in the mediastinum consistent with esophageal perforation (Figure 2). One day after admission a Hypaque esophagram showed trauma to the posterior cervical region of the esophagus, but no leak of contrast material into the mediastinum. The patient was managed conservatively with intravenous (IV) antibiotics and nothing by mouth. Stools were screened and 48 hours after admission the patient (painlessly) passed a piece of glass with a very sharp point (Figure 3) correlating in size and shape to the foreign body seen on the previous abdominal radiograph. The glass had apparently fallen into the patient's restaurant meal the night of admission. The patient did well and was discharged 6 days after admission. He asked the manager of his favorite restaurant to reimburse him the $200 copayment required for hospitalization required under his Preferred Provider Plan; the request was immediately honored.

Figure 1

Figure 2

Figure 3

Esophageal perforation is an emergency because of its high mortality rate. The most frequent cause is iatrogenic with instrumentation from endoscopic procedures. Other causes include foreign body ingestion (as in this case), trauma, operative injury, and tumor.1 Aggressive surgical intervention vs. conservative nonsurgical management remains a controversial topic.2 Early‐contained perforations can be managed successfully by limiting oral intake and giving parenteral antibiotics.3, 4 Any signs of sepsis, deterioration in the patient's condition, or uncontained rupture warrants immediate surgical intervention.14

A 33‐year‐old African American man was seen in the emergency department for bilateral wrist pain and forearm swelling. He was a professional mover and had transported a piano 3 days before. Several hours after the move, he noticed wrist pain and a few small red, slightly pruritic bumps on the palmar aspect of both wrists. The following day, he developed nausea, vomiting, and watery diarrhea. His wrists and forearms became more swollen and painful.

The most distinctive aspect of the patient's symptom complex is forearm swelling. I am not certain if the primary pathology lies in the wrist or in the forearm. Examination should focus on the presence or absence of arthritis and whether the forearm swelling is simply adjacent to the wrist and rash or extends the entire length from the wrist to the elbow.

Strenuous lifting can lead to rhabdomyolysis and forearm compartment syndrome. However, transporting a piano is not an unusual task for a professional mover. More common causes of arm swelling such as fracture, cellulitis, deep vein thrombosis, and lymphatic obstruction are possible but are unexpected in this patient because of the bilateral findings. An inflammatory myopathy would not present in the distal extremities, but an infectious myositis, such as trichinosis (which can have early gastrointestinal symptoms), could.

Given the patient's age and the temporal correlation, I am inclined to pursue a unifying diagnosis between his gastrointestinal and upper extremity symptoms. The gastrointestinal symptoms could reflect vasculitis (eg, polyarteritis nodosa) or a nonspecific manifestation of systemic illness (eg, sepsis), whereas the rash could be due to infection with petechiae (eg, meningococcemia, gonococcemia, or endocarditis) or an infection with a predilection for peripheral skin lesions that progress centripetally as the illness progresses, such as Rocky Mountain spotted fever.

Although the patient had not seen spiders, he was concerned that his skin lesions might have resulted from spider bites. He had a history of atopic dermatitis but took no medications, did not smoke, and drank alcohol rarely. He lived in Denver, CO, had not traveled outside of the region, and had not visited rural areas. He was monogamous with a female partner and had no known exposures to human immunodeficiency virus. He had no family history of rheumatological disorders.

Patients and physicians frequently attribute papules, ulcers, or necrotic skin lesions to spider bites far out of proportion to their true prevalence. Bites by more innocuous arthropods such as ticks, fleas, bedbugs, or mites are much more common. The symmetry of the lesions and extensive swelling, however, make such bites unlikely.

Although the patient hails from the southwestern United States, there is no compelling evidence for endemic illness. Hantavirus infection causes a viral prodrome, but a severe pulmonary syndrome is its primary manifestation. Plague presents in bubonic, pulmonary, or septicemic forms, but other than the gastrointestinal symptoms, there is nothing to suggest such a systemic illness. The initial pulmonary infection of coccidioidomycosis is often unnoticed, and patients may present with extrapulmonary manifestations, including skin lesions and skeletal disease.

More common ailments remain on the differential. Disseminated gonococcemia must be considered in a sexually active adult with skin lesions and what may be tenosynovitis or arthritis of the distal extremities. The history of atopic dermatitis supports a diagnosis of allergic or contact dermatitis on the wrists (perhaps inoculated during the move), explaining the rash and adjacent swelling (but not the gastrointestinal symptoms).

The patient's temperature was 36.5C with a pulse of 103 beats per minute, a blood pressure of 103/67 mm Hg, and a respiratory rate of 24 breaths per minute. He appeared uncomfortable and was in moderate distress. His sclerae were injected, and his mucous membranes were dry. He had diffusely swollen fingers and firm nonpitting edema in both hands and forearms. His wrists and hands were tender and warm but had no appreciable redness. Two 1‐mm ulcerated lesions were present on his right wrist, and one 2‐mm ulcerated lesion was present on his left wrist. No lymphadenopathy was present.

The patient meets the criteria for systemic inflammatory response syndrome (SIRS), and considering his ill appearance, I am concerned that he may have an infectious process that has evolved into sepsis. The absence of cutaneous erythema rules out cellulitis, and although skin findings in necrotizing fasciitis can be modest in comparison with the underlying infection, an examination with nothing more than punctate ulcerations would be atypical. Plague and tularemia cause ulcerative skin lesions and systemic illness but usually have prominent lymphadenopathy. Cutaneous anthrax can cause intense local edema but is usually accompanied by some degree of necrosis.

I am struck by the degree of local edema. It would be a remarkable coincidence for spider bites to occur on both wrists; however, given the size of the lesions, this remains a consideration. Spider bites can cause severe local swelling and occasionally even SIRS.

The white blood cellcount was 6000/mm³ with a normal differential. The hemoglobin level was 16.6 g/dL, and the platelet count was 227,000/mm³. The serum sodium level was measured to be 135 mmol/L, the potassium level was 3.6 mmol/L, the chloride level was 94 mmol/L, and the bicarbonate level was 9 mmol/L. The urea nitrogen level was measured to be 48 mg/dL (normal, 622), and the serum creatinine level was 5.6 mg/dL (normal, 0.41.2). An arterial blood gas test on room air revealed a pH of 7.22, a partial pressure of carbon dioxide of 24 mm Hg, and a partial pressure of oxygen of 128 mm Hg. Liver enzymes were normal. The serum creatine kinase level was measured to be 194 U/L (normal, 0250). A chest radiograph revealed clear lung fields.

The anion gap is elevated and could be explained in part by renal failure, but it is quite pronounced, and I suspect that there is lactic acidosis either from SIRS and systemic hypoperfusion or from local underperfusion of the distal upper extremities (eg, compartment syndrome). Rhabdomyolysis can cause an anion gap acidosis and acute renal failure but is ruled out by a normal creatine kinase level.

Rheumatological diseases (such as polyarteritis nodosa and systemic scleroderma) can cause renal failure, gastrointestinal symptoms, and cutaneous lesions with skin ulceration but are often accompanied by hypertension. The combination of SIRS and volume depletion from nausea, vomiting, and diarrhea seems more likely, although an etiology for SIRS remains elusive. I suspect infectionmost likely of the upper extremitiesis the underlying cause in this patient despite his normal body temperature and white cell count. I would therefore start with plain films of the arms and hands. If these are unrevealing, I would proceed with an ultrasound of the arms to evaluate the soft tissues and particularly the vessels. Although imaging studies are unlikely to provide a diagnosis, they will provide guidance in choosing the next step, such as biopsy or culture.

The patient wasvolume‐resuscitated with saline and treated with clindamycin, pipercillin‐tazobactam, and vancomycin. Radiographs of the wrists and hands demonstrated edema but no subcutaneous gas and no abnormalities of the bones or joints. An ultrasound of the upper extremities was negative for superficial or deep venous thrombosis. Renal ultrasonography revealed normal‐sized kidneys and no hydronephrosis.

Short of superior vena cava thrombosis, which the ultrasound could not visualize, deep venous thrombosis can be ruled out with confidence. Superior vena cava syndrome could account for the bilateral upper extremity symptoms, but complete sparing of the face would be unusual, and the chest radiograph was normal.

Despite the bilateral nature, I doubt that this patient has an arthritis of the wrists and hands (eg, rheumatoid or psoriatic arthritis). The plain films did not show evidence of joint destruction, although that would not be expected in the first few days of most noninfectious arthropathies. Many rheumatological diseases and vasculitides have renal manifestations, but I do not find convincing evidence of such diseases yet.

Despite intravenous fluidsover the next 4 hours, the patient remained anuric. His upper extremity edema worsened, and he became increasingly tachycardic and hypotensive. Vasopressor support was required. Repeat laboratory studies demonstrated a serum creatinine level of 7.0 mg/dL, a bicarbonate level of 6 mmol/L, and a creatine kinase level of 1183 U/L. The serum lactate level was measured to be 6.5. Blood cultures obtained on admission were negative. An echocardiogram demonstrated marked impairment of left ventricular systolic dysfunction with an estimated ejection fraction of 35%. Repeat chest radiography revealed only mild pulmonary vascular congestion.

The patient has become progressively ill despite initial resuscitation. Any infection (cellulitis, fasciitis, or myositis) may progress to such a point (although to do so without fever or leukocytosis with this degree of illness is unusual) and would prompt me to ask a surgeon to explore the upper extremities for diagnostic and possibly therapeutic purposes. Among the spectrum of possible soft tissue infections, this clinical presentation is most consistent with pyomyositis (caused by Staphylococcus aureus) because of the relatively modest creatine kinase elevations that accompany it and the overall absence of cutaneous findings (save the punctuate lesions), which I would expect to be present with cellulitis or necrotizing fasciitis. A deep forearm infection and perhaps compartment syndrome leading to sepsis could explain lactic acidosis, decreased cardiac function, hypotension, and acute renal failure.

Because of the unusual characteristics noted so farparticularly the bilateral diseaseI continue to also consider systemic diseases that can cause skin lesions, cardiomyopathy, renal disease, ocular involvement (eg, keratoconjunctivitis or uveitis), and myositis. Sarcoidosis is possible, although in most of the aforementioned organs, histological disease is far more common than clinical disease, and sarcoidosis typically does not cause this degree of illness. Furthermore, over 90% of patients with sarcoidosis have pulmonary involvement. Polyarteritis nodosa could explain the multiorgan involvement and the brisk pace. If no infection is present on exploration, I would ask the surgeon to biopsy the muscle, particularly looking for granulomas or vasculitis. A progressive soft tissue infection leading to sepsis remains my leading consideration at this point.

Surgical consultation was obtained. A muscle biopsy of the left biceps and left forearm revealed group A streptococcus within the muscle and evidence of necrosis (Figure 1). Debridement of both arms and wrists was performed. The patient subsequently developed erythema of the palms and soles followed by diffuse sloughing of the skin. Streptococcal toxic shock syndrome with necrotizing myositis was diagnosed, and the antimicrobial regimen was changed to intravenous penicillin, clindamycin, and intravenous immunoglobulin G. Repeat debridement of the arms was required.

Figure 1

Even when a soft tissue infection is suspected, it can be challenging to preoperatively localize or characterize with precision. In retrospect, the overall severity of his illness and his previously good health status perhaps favor necrotizing myositis (and fasciitis) over pyomyositis.

I may have put undue emphasis on the absence of skin findings. Although the symmetric nature of the disease is unusual, the small skin lesions may have been portals of entry, and in retrospect, they represent the tip of the iceberg. The absence of fever and leukocytosisor hypothermia and leukopeniain a young, previously healthy patient along with the bilateral and symmetric findings made me hesitant to definitely label his illness as a deep soft tissue infection early on, but the gravity of the illness, the lack of a plausible alternative explanation, and his precipitous decline all made surgical exploration imperative.

The patient's skin sloughing progressed, and he was transferred to a regional burn unit. Three additional operations were required for debridement of both upper extremities. Despite apparent control of the initial infection, the patient continued to require significant hemodynamic and ventilator support. He subsequently developed neutropenia, thrombocytopenia, and Escherichia coli urosepsis. Necrosis of the lower extremities developed, and additional surgical debridement was recommended. After extensive discussion regarding prognosis, the family decided against further surgery and withdrew life support. The patient died shortly after extubation. The family refused an autopsy.

Commentary

Expert clinicians employ a variety of approaches to solve complex clinical problems. One of the most effective strategies is pattern recognition, in which the clinician divides the case into recognizable portions and compares these to previous cases that he or she has encountered.1, 2 If patterns from the new case appear similar or identical to those of previous cases, a diagnosis can be made quickly and without the need for unnecessary testing. However, when features of the case are unusual or atypical, pattern recognition may be disrupted. For example, although the discussant suspected a soft tissue infection, a number of features (including a normal white blood cell count, minimal skin findings, and a bilateral and symmetric distribution of swelling) did not match his illness script (a mental representation of a disease) of necrotizing fasciitis.

When pattern recognition fails, other strategies are available.3 Hypothetico‐deductive reasoning is a data‐to‐diagnosis method whereby the clinician uses the presenting information to construct a list of diagnostic possibilities.4 Additional testing and gathering of information are then used to continuously revise the diagnostic possibilities until confirmatory information is obtained and the diagnosis is established. After a pattern failed to materialize, the discussant employed this analytical strategy by noting the unusual characteristics of the case and incorporating laboratory and physiological data to revise his differential diagnosis. As a result, he requested the appropriate diagnostic test: surgical exploration of the forearms.

Necrotizing soft tissue infections are characterized by fulminant tissue destruction, rapid spread along tissue planes, and local vascular thrombosis. Mixed aerobic and anaerobic infections typically occur after penetrating skin injury or following surgery in patients with diabetes mellitus or vascular disease. In contrast, monomicrobial infections with S. aureus or group A streptococcus generally occur in healthy individuals. The prevalence of necrotizing group A streptococcal infections has increased dramatically in the last 15 to 20 years.5 Over one‐third of these cases are complicated by toxic shock syndrome.57 Mortality rates for necrotizing fasciitis with toxic shock exceed 30%, and early surgical consultation is directly associated with a reduction in morbidity and mortality.5, 8

Toxic shock is an inflammatory response syndrome caused by release of exotoxins from group A streptococcus and S. aureus.9, 10 In streptococcal toxic shock, Streptococcus pyogenes exotoxin A and Streptococcus pyogenes exotoxin B are the major toxins produced.10 These toxins activate the systemic production of inflammatory cytokines such as interleukin‐1, gamma‐interferon, and tumor necrosis factor, resulting in capillary leak, systemic hypotension, tissue hypoperfusion, and organ failure. The most common initial symptom is diffuse or localized pain that is severe and abrupt in onset and often precedes or is out of proportion to other physical findings of soft tissue infection.8 Up to 20% of individuals may also develop a viral‐like syndrome with myalgias, fever, nausea, vomiting, and diarrhea.11 Erythroderma of the skin and mucous membranes can be another early finding. The rash is diffuse, erythematous, and macular, resembling a sunburn. It involves the palms and soles but can be subtle and fleeting. Erythroderma may be particularly difficult to detect in dark‐skinned individuals.12 It is also important to consider that the absence of fever, erythroderma, or leukocytosis does not necessarily rule out the possibility of serious infection in necrotizing fasciitis patients, as the development of these signs and symptoms may occur later in the disease process.

The most common portals of entry for group A streptococcus are the skin, vagina, and pharynx. Predisposing factors include varicella infection, penetrating injuries, minor cuts, burns, splinters, and surgery. Interestingly, a portal of entry cannot be identified in 45% of cases.8 These patients in particular are at risk for developing severe necrotizing myositis or fasciitis at the site of a minor injury such as a strained muscle. Hematogenous translocation from the pharynx to the site of injury is the probable mechanism13 and would provide one scenario by which our piano mover developed bilateral and symmetric disease. An alternative explanation would be direct extension of bacteria from small breaks in the skin to adjacent areas of muscle strain. Regardless of the portal of entry, as the small skin lesions demonstrate in this patient, the smallest physical finding can represent the tip of the iceberg.

A 33‐year‐old African American man was seen in the emergency department for bilateral wrist pain and forearm swelling. He was a professional mover and had transported a piano 3 days before. Several hours after the move, he noticed wrist pain and a few small red, slightly pruritic bumps on the palmar aspect of both wrists. The following day, he developed nausea, vomiting, and watery diarrhea. His wrists and forearms became more swollen and painful.

The most distinctive aspect of the patient's symptom complex is forearm swelling. I am not certain if the primary pathology lies in the wrist or in the forearm. Examination should focus on the presence or absence of arthritis and whether the forearm swelling is simply adjacent to the wrist and rash or extends the entire length from the wrist to the elbow.

Strenuous lifting can lead to rhabdomyolysis and forearm compartment syndrome. However, transporting a piano is not an unusual task for a professional mover. More common causes of arm swelling such as fracture, cellulitis, deep vein thrombosis, and lymphatic obstruction are possible but are unexpected in this patient because of the bilateral findings. An inflammatory myopathy would not present in the distal extremities, but an infectious myositis, such as trichinosis (which can have early gastrointestinal symptoms), could.

Given the patient's age and the temporal correlation, I am inclined to pursue a unifying diagnosis between his gastrointestinal and upper extremity symptoms. The gastrointestinal symptoms could reflect vasculitis (eg, polyarteritis nodosa) or a nonspecific manifestation of systemic illness (eg, sepsis), whereas the rash could be due to infection with petechiae (eg, meningococcemia, gonococcemia, or endocarditis) or an infection with a predilection for peripheral skin lesions that progress centripetally as the illness progresses, such as Rocky Mountain spotted fever.

Although the patient had not seen spiders, he was concerned that his skin lesions might have resulted from spider bites. He had a history of atopic dermatitis but took no medications, did not smoke, and drank alcohol rarely. He lived in Denver, CO, had not traveled outside of the region, and had not visited rural areas. He was monogamous with a female partner and had no known exposures to human immunodeficiency virus. He had no family history of rheumatological disorders.

Patients and physicians frequently attribute papules, ulcers, or necrotic skin lesions to spider bites far out of proportion to their true prevalence. Bites by more innocuous arthropods such as ticks, fleas, bedbugs, or mites are much more common. The symmetry of the lesions and extensive swelling, however, make such bites unlikely.

Although the patient hails from the southwestern United States, there is no compelling evidence for endemic illness. Hantavirus infection causes a viral prodrome, but a severe pulmonary syndrome is its primary manifestation. Plague presents in bubonic, pulmonary, or septicemic forms, but other than the gastrointestinal symptoms, there is nothing to suggest such a systemic illness. The initial pulmonary infection of coccidioidomycosis is often unnoticed, and patients may present with extrapulmonary manifestations, including skin lesions and skeletal disease.

More common ailments remain on the differential. Disseminated gonococcemia must be considered in a sexually active adult with skin lesions and what may be tenosynovitis or arthritis of the distal extremities. The history of atopic dermatitis supports a diagnosis of allergic or contact dermatitis on the wrists (perhaps inoculated during the move), explaining the rash and adjacent swelling (but not the gastrointestinal symptoms).

The patient's temperature was 36.5C with a pulse of 103 beats per minute, a blood pressure of 103/67 mm Hg, and a respiratory rate of 24 breaths per minute. He appeared uncomfortable and was in moderate distress. His sclerae were injected, and his mucous membranes were dry. He had diffusely swollen fingers and firm nonpitting edema in both hands and forearms. His wrists and hands were tender and warm but had no appreciable redness. Two 1‐mm ulcerated lesions were present on his right wrist, and one 2‐mm ulcerated lesion was present on his left wrist. No lymphadenopathy was present.

The patient meets the criteria for systemic inflammatory response syndrome (SIRS), and considering his ill appearance, I am concerned that he may have an infectious process that has evolved into sepsis. The absence of cutaneous erythema rules out cellulitis, and although skin findings in necrotizing fasciitis can be modest in comparison with the underlying infection, an examination with nothing more than punctate ulcerations would be atypical. Plague and tularemia cause ulcerative skin lesions and systemic illness but usually have prominent lymphadenopathy. Cutaneous anthrax can cause intense local edema but is usually accompanied by some degree of necrosis.

I am struck by the degree of local edema. It would be a remarkable coincidence for spider bites to occur on both wrists; however, given the size of the lesions, this remains a consideration. Spider bites can cause severe local swelling and occasionally even SIRS.

The white blood cellcount was 6000/mm³ with a normal differential. The hemoglobin level was 16.6 g/dL, and the platelet count was 227,000/mm³. The serum sodium level was measured to be 135 mmol/L, the potassium level was 3.6 mmol/L, the chloride level was 94 mmol/L, and the bicarbonate level was 9 mmol/L. The urea nitrogen level was measured to be 48 mg/dL (normal, 622), and the serum creatinine level was 5.6 mg/dL (normal, 0.41.2). An arterial blood gas test on room air revealed a pH of 7.22, a partial pressure of carbon dioxide of 24 mm Hg, and a partial pressure of oxygen of 128 mm Hg. Liver enzymes were normal. The serum creatine kinase level was measured to be 194 U/L (normal, 0250). A chest radiograph revealed clear lung fields.

The anion gap is elevated and could be explained in part by renal failure, but it is quite pronounced, and I suspect that there is lactic acidosis either from SIRS and systemic hypoperfusion or from local underperfusion of the distal upper extremities (eg, compartment syndrome). Rhabdomyolysis can cause an anion gap acidosis and acute renal failure but is ruled out by a normal creatine kinase level.

Rheumatological diseases (such as polyarteritis nodosa and systemic scleroderma) can cause renal failure, gastrointestinal symptoms, and cutaneous lesions with skin ulceration but are often accompanied by hypertension. The combination of SIRS and volume depletion from nausea, vomiting, and diarrhea seems more likely, although an etiology for SIRS remains elusive. I suspect infectionmost likely of the upper extremitiesis the underlying cause in this patient despite his normal body temperature and white cell count. I would therefore start with plain films of the arms and hands. If these are unrevealing, I would proceed with an ultrasound of the arms to evaluate the soft tissues and particularly the vessels. Although imaging studies are unlikely to provide a diagnosis, they will provide guidance in choosing the next step, such as biopsy or culture.

The patient wasvolume‐resuscitated with saline and treated with clindamycin, pipercillin‐tazobactam, and vancomycin. Radiographs of the wrists and hands demonstrated edema but no subcutaneous gas and no abnormalities of the bones or joints. An ultrasound of the upper extremities was negative for superficial or deep venous thrombosis. Renal ultrasonography revealed normal‐sized kidneys and no hydronephrosis.

Short of superior vena cava thrombosis, which the ultrasound could not visualize, deep venous thrombosis can be ruled out with confidence. Superior vena cava syndrome could account for the bilateral upper extremity symptoms, but complete sparing of the face would be unusual, and the chest radiograph was normal.

Despite the bilateral nature, I doubt that this patient has an arthritis of the wrists and hands (eg, rheumatoid or psoriatic arthritis). The plain films did not show evidence of joint destruction, although that would not be expected in the first few days of most noninfectious arthropathies. Many rheumatological diseases and vasculitides have renal manifestations, but I do not find convincing evidence of such diseases yet.

Despite intravenous fluidsover the next 4 hours, the patient remained anuric. His upper extremity edema worsened, and he became increasingly tachycardic and hypotensive. Vasopressor support was required. Repeat laboratory studies demonstrated a serum creatinine level of 7.0 mg/dL, a bicarbonate level of 6 mmol/L, and a creatine kinase level of 1183 U/L. The serum lactate level was measured to be 6.5. Blood cultures obtained on admission were negative. An echocardiogram demonstrated marked impairment of left ventricular systolic dysfunction with an estimated ejection fraction of 35%. Repeat chest radiography revealed only mild pulmonary vascular congestion.

The patient has become progressively ill despite initial resuscitation. Any infection (cellulitis, fasciitis, or myositis) may progress to such a point (although to do so without fever or leukocytosis with this degree of illness is unusual) and would prompt me to ask a surgeon to explore the upper extremities for diagnostic and possibly therapeutic purposes. Among the spectrum of possible soft tissue infections, this clinical presentation is most consistent with pyomyositis (caused by Staphylococcus aureus) because of the relatively modest creatine kinase elevations that accompany it and the overall absence of cutaneous findings (save the punctuate lesions), which I would expect to be present with cellulitis or necrotizing fasciitis. A deep forearm infection and perhaps compartment syndrome leading to sepsis could explain lactic acidosis, decreased cardiac function, hypotension, and acute renal failure.

Because of the unusual characteristics noted so farparticularly the bilateral diseaseI continue to also consider systemic diseases that can cause skin lesions, cardiomyopathy, renal disease, ocular involvement (eg, keratoconjunctivitis or uveitis), and myositis. Sarcoidosis is possible, although in most of the aforementioned organs, histological disease is far more common than clinical disease, and sarcoidosis typically does not cause this degree of illness. Furthermore, over 90% of patients with sarcoidosis have pulmonary involvement. Polyarteritis nodosa could explain the multiorgan involvement and the brisk pace. If no infection is present on exploration, I would ask the surgeon to biopsy the muscle, particularly looking for granulomas or vasculitis. A progressive soft tissue infection leading to sepsis remains my leading consideration at this point.

Surgical consultation was obtained. A muscle biopsy of the left biceps and left forearm revealed group A streptococcus within the muscle and evidence of necrosis (Figure 1). Debridement of both arms and wrists was performed. The patient subsequently developed erythema of the palms and soles followed by diffuse sloughing of the skin. Streptococcal toxic shock syndrome with necrotizing myositis was diagnosed, and the antimicrobial regimen was changed to intravenous penicillin, clindamycin, and intravenous immunoglobulin G. Repeat debridement of the arms was required.

Figure 1

Even when a soft tissue infection is suspected, it can be challenging to preoperatively localize or characterize with precision. In retrospect, the overall severity of his illness and his previously good health status perhaps favor necrotizing myositis (and fasciitis) over pyomyositis.

I may have put undue emphasis on the absence of skin findings. Although the symmetric nature of the disease is unusual, the small skin lesions may have been portals of entry, and in retrospect, they represent the tip of the iceberg. The absence of fever and leukocytosisor hypothermia and leukopeniain a young, previously healthy patient along with the bilateral and symmetric findings made me hesitant to definitely label his illness as a deep soft tissue infection early on, but the gravity of the illness, the lack of a plausible alternative explanation, and his precipitous decline all made surgical exploration imperative.

The patient's skin sloughing progressed, and he was transferred to a regional burn unit. Three additional operations were required for debridement of both upper extremities. Despite apparent control of the initial infection, the patient continued to require significant hemodynamic and ventilator support. He subsequently developed neutropenia, thrombocytopenia, and Escherichia coli urosepsis. Necrosis of the lower extremities developed, and additional surgical debridement was recommended. After extensive discussion regarding prognosis, the family decided against further surgery and withdrew life support. The patient died shortly after extubation. The family refused an autopsy.

Commentary

Expert clinicians employ a variety of approaches to solve complex clinical problems. One of the most effective strategies is pattern recognition, in which the clinician divides the case into recognizable portions and compares these to previous cases that he or she has encountered.1, 2 If patterns from the new case appear similar or identical to those of previous cases, a diagnosis can be made quickly and without the need for unnecessary testing. However, when features of the case are unusual or atypical, pattern recognition may be disrupted. For example, although the discussant suspected a soft tissue infection, a number of features (including a normal white blood cell count, minimal skin findings, and a bilateral and symmetric distribution of swelling) did not match his illness script (a mental representation of a disease) of necrotizing fasciitis.

When pattern recognition fails, other strategies are available.3 Hypothetico‐deductive reasoning is a data‐to‐diagnosis method whereby the clinician uses the presenting information to construct a list of diagnostic possibilities.4 Additional testing and gathering of information are then used to continuously revise the diagnostic possibilities until confirmatory information is obtained and the diagnosis is established. After a pattern failed to materialize, the discussant employed this analytical strategy by noting the unusual characteristics of the case and incorporating laboratory and physiological data to revise his differential diagnosis. As a result, he requested the appropriate diagnostic test: surgical exploration of the forearms.

Necrotizing soft tissue infections are characterized by fulminant tissue destruction, rapid spread along tissue planes, and local vascular thrombosis. Mixed aerobic and anaerobic infections typically occur after penetrating skin injury or following surgery in patients with diabetes mellitus or vascular disease. In contrast, monomicrobial infections with S. aureus or group A streptococcus generally occur in healthy individuals. The prevalence of necrotizing group A streptococcal infections has increased dramatically in the last 15 to 20 years.5 Over one‐third of these cases are complicated by toxic shock syndrome.57 Mortality rates for necrotizing fasciitis with toxic shock exceed 30%, and early surgical consultation is directly associated with a reduction in morbidity and mortality.5, 8

Toxic shock is an inflammatory response syndrome caused by release of exotoxins from group A streptococcus and S. aureus.9, 10 In streptococcal toxic shock, Streptococcus pyogenes exotoxin A and Streptococcus pyogenes exotoxin B are the major toxins produced.10 These toxins activate the systemic production of inflammatory cytokines such as interleukin‐1, gamma‐interferon, and tumor necrosis factor, resulting in capillary leak, systemic hypotension, tissue hypoperfusion, and organ failure. The most common initial symptom is diffuse or localized pain that is severe and abrupt in onset and often precedes or is out of proportion to other physical findings of soft tissue infection.8 Up to 20% of individuals may also develop a viral‐like syndrome with myalgias, fever, nausea, vomiting, and diarrhea.11 Erythroderma of the skin and mucous membranes can be another early finding. The rash is diffuse, erythematous, and macular, resembling a sunburn. It involves the palms and soles but can be subtle and fleeting. Erythroderma may be particularly difficult to detect in dark‐skinned individuals.12 It is also important to consider that the absence of fever, erythroderma, or leukocytosis does not necessarily rule out the possibility of serious infection in necrotizing fasciitis patients, as the development of these signs and symptoms may occur later in the disease process.

The most common portals of entry for group A streptococcus are the skin, vagina, and pharynx. Predisposing factors include varicella infection, penetrating injuries, minor cuts, burns, splinters, and surgery. Interestingly, a portal of entry cannot be identified in 45% of cases.8 These patients in particular are at risk for developing severe necrotizing myositis or fasciitis at the site of a minor injury such as a strained muscle. Hematogenous translocation from the pharynx to the site of injury is the probable mechanism13 and would provide one scenario by which our piano mover developed bilateral and symmetric disease. An alternative explanation would be direct extension of bacteria from small breaks in the skin to adjacent areas of muscle strain. Regardless of the portal of entry, as the small skin lesions demonstrate in this patient, the smallest physical finding can represent the tip of the iceberg.

A 33‐year‐old African American man was seen in the emergency department for bilateral wrist pain and forearm swelling. He was a professional mover and had transported a piano 3 days before. Several hours after the move, he noticed wrist pain and a few small red, slightly pruritic bumps on the palmar aspect of both wrists. The following day, he developed nausea, vomiting, and watery diarrhea. His wrists and forearms became more swollen and painful.

The most distinctive aspect of the patient's symptom complex is forearm swelling. I am not certain if the primary pathology lies in the wrist or in the forearm. Examination should focus on the presence or absence of arthritis and whether the forearm swelling is simply adjacent to the wrist and rash or extends the entire length from the wrist to the elbow.

Strenuous lifting can lead to rhabdomyolysis and forearm compartment syndrome. However, transporting a piano is not an unusual task for a professional mover. More common causes of arm swelling such as fracture, cellulitis, deep vein thrombosis, and lymphatic obstruction are possible but are unexpected in this patient because of the bilateral findings. An inflammatory myopathy would not present in the distal extremities, but an infectious myositis, such as trichinosis (which can have early gastrointestinal symptoms), could.

Given the patient's age and the temporal correlation, I am inclined to pursue a unifying diagnosis between his gastrointestinal and upper extremity symptoms. The gastrointestinal symptoms could reflect vasculitis (eg, polyarteritis nodosa) or a nonspecific manifestation of systemic illness (eg, sepsis), whereas the rash could be due to infection with petechiae (eg, meningococcemia, gonococcemia, or endocarditis) or an infection with a predilection for peripheral skin lesions that progress centripetally as the illness progresses, such as Rocky Mountain spotted fever.

Although the patient had not seen spiders, he was concerned that his skin lesions might have resulted from spider bites. He had a history of atopic dermatitis but took no medications, did not smoke, and drank alcohol rarely. He lived in Denver, CO, had not traveled outside of the region, and had not visited rural areas. He was monogamous with a female partner and had no known exposures to human immunodeficiency virus. He had no family history of rheumatological disorders.

Patients and physicians frequently attribute papules, ulcers, or necrotic skin lesions to spider bites far out of proportion to their true prevalence. Bites by more innocuous arthropods such as ticks, fleas, bedbugs, or mites are much more common. The symmetry of the lesions and extensive swelling, however, make such bites unlikely.

Although the patient hails from the southwestern United States, there is no compelling evidence for endemic illness. Hantavirus infection causes a viral prodrome, but a severe pulmonary syndrome is its primary manifestation. Plague presents in bubonic, pulmonary, or septicemic forms, but other than the gastrointestinal symptoms, there is nothing to suggest such a systemic illness. The initial pulmonary infection of coccidioidomycosis is often unnoticed, and patients may present with extrapulmonary manifestations, including skin lesions and skeletal disease.

More common ailments remain on the differential. Disseminated gonococcemia must be considered in a sexually active adult with skin lesions and what may be tenosynovitis or arthritis of the distal extremities. The history of atopic dermatitis supports a diagnosis of allergic or contact dermatitis on the wrists (perhaps inoculated during the move), explaining the rash and adjacent swelling (but not the gastrointestinal symptoms).

The patient's temperature was 36.5C with a pulse of 103 beats per minute, a blood pressure of 103/67 mm Hg, and a respiratory rate of 24 breaths per minute. He appeared uncomfortable and was in moderate distress. His sclerae were injected, and his mucous membranes were dry. He had diffusely swollen fingers and firm nonpitting edema in both hands and forearms. His wrists and hands were tender and warm but had no appreciable redness. Two 1‐mm ulcerated lesions were present on his right wrist, and one 2‐mm ulcerated lesion was present on his left wrist. No lymphadenopathy was present.

The patient meets the criteria for systemic inflammatory response syndrome (SIRS), and considering his ill appearance, I am concerned that he may have an infectious process that has evolved into sepsis. The absence of cutaneous erythema rules out cellulitis, and although skin findings in necrotizing fasciitis can be modest in comparison with the underlying infection, an examination with nothing more than punctate ulcerations would be atypical. Plague and tularemia cause ulcerative skin lesions and systemic illness but usually have prominent lymphadenopathy. Cutaneous anthrax can cause intense local edema but is usually accompanied by some degree of necrosis.

I am struck by the degree of local edema. It would be a remarkable coincidence for spider bites to occur on both wrists; however, given the size of the lesions, this remains a consideration. Spider bites can cause severe local swelling and occasionally even SIRS.

The white blood cellcount was 6000/mm³ with a normal differential. The hemoglobin level was 16.6 g/dL, and the platelet count was 227,000/mm³. The serum sodium level was measured to be 135 mmol/L, the potassium level was 3.6 mmol/L, the chloride level was 94 mmol/L, and the bicarbonate level was 9 mmol/L. The urea nitrogen level was measured to be 48 mg/dL (normal, 622), and the serum creatinine level was 5.6 mg/dL (normal, 0.41.2). An arterial blood gas test on room air revealed a pH of 7.22, a partial pressure of carbon dioxide of 24 mm Hg, and a partial pressure of oxygen of 128 mm Hg. Liver enzymes were normal. The serum creatine kinase level was measured to be 194 U/L (normal, 0250). A chest radiograph revealed clear lung fields.

The anion gap is elevated and could be explained in part by renal failure, but it is quite pronounced, and I suspect that there is lactic acidosis either from SIRS and systemic hypoperfusion or from local underperfusion of the distal upper extremities (eg, compartment syndrome). Rhabdomyolysis can cause an anion gap acidosis and acute renal failure but is ruled out by a normal creatine kinase level.

Rheumatological diseases (such as polyarteritis nodosa and systemic scleroderma) can cause renal failure, gastrointestinal symptoms, and cutaneous lesions with skin ulceration but are often accompanied by hypertension. The combination of SIRS and volume depletion from nausea, vomiting, and diarrhea seems more likely, although an etiology for SIRS remains elusive. I suspect infectionmost likely of the upper extremitiesis the underlying cause in this patient despite his normal body temperature and white cell count. I would therefore start with plain films of the arms and hands. If these are unrevealing, I would proceed with an ultrasound of the arms to evaluate the soft tissues and particularly the vessels. Although imaging studies are unlikely to provide a diagnosis, they will provide guidance in choosing the next step, such as biopsy or culture.

The patient wasvolume‐resuscitated with saline and treated with clindamycin, pipercillin‐tazobactam, and vancomycin. Radiographs of the wrists and hands demonstrated edema but no subcutaneous gas and no abnormalities of the bones or joints. An ultrasound of the upper extremities was negative for superficial or deep venous thrombosis. Renal ultrasonography revealed normal‐sized kidneys and no hydronephrosis.

Short of superior vena cava thrombosis, which the ultrasound could not visualize, deep venous thrombosis can be ruled out with confidence. Superior vena cava syndrome could account for the bilateral upper extremity symptoms, but complete sparing of the face would be unusual, and the chest radiograph was normal.

Despite the bilateral nature, I doubt that this patient has an arthritis of the wrists and hands (eg, rheumatoid or psoriatic arthritis). The plain films did not show evidence of joint destruction, although that would not be expected in the first few days of most noninfectious arthropathies. Many rheumatological diseases and vasculitides have renal manifestations, but I do not find convincing evidence of such diseases yet.

Despite intravenous fluidsover the next 4 hours, the patient remained anuric. His upper extremity edema worsened, and he became increasingly tachycardic and hypotensive. Vasopressor support was required. Repeat laboratory studies demonstrated a serum creatinine level of 7.0 mg/dL, a bicarbonate level of 6 mmol/L, and a creatine kinase level of 1183 U/L. The serum lactate level was measured to be 6.5. Blood cultures obtained on admission were negative. An echocardiogram demonstrated marked impairment of left ventricular systolic dysfunction with an estimated ejection fraction of 35%. Repeat chest radiography revealed only mild pulmonary vascular congestion.

The patient has become progressively ill despite initial resuscitation. Any infection (cellulitis, fasciitis, or myositis) may progress to such a point (although to do so without fever or leukocytosis with this degree of illness is unusual) and would prompt me to ask a surgeon to explore the upper extremities for diagnostic and possibly therapeutic purposes. Among the spectrum of possible soft tissue infections, this clinical presentation is most consistent with pyomyositis (caused by Staphylococcus aureus) because of the relatively modest creatine kinase elevations that accompany it and the overall absence of cutaneous findings (save the punctuate lesions), which I would expect to be present with cellulitis or necrotizing fasciitis. A deep forearm infection and perhaps compartment syndrome leading to sepsis could explain lactic acidosis, decreased cardiac function, hypotension, and acute renal failure.

Because of the unusual characteristics noted so farparticularly the bilateral diseaseI continue to also consider systemic diseases that can cause skin lesions, cardiomyopathy, renal disease, ocular involvement (eg, keratoconjunctivitis or uveitis), and myositis. Sarcoidosis is possible, although in most of the aforementioned organs, histological disease is far more common than clinical disease, and sarcoidosis typically does not cause this degree of illness. Furthermore, over 90% of patients with sarcoidosis have pulmonary involvement. Polyarteritis nodosa could explain the multiorgan involvement and the brisk pace. If no infection is present on exploration, I would ask the surgeon to biopsy the muscle, particularly looking for granulomas or vasculitis. A progressive soft tissue infection leading to sepsis remains my leading consideration at this point.

Surgical consultation was obtained. A muscle biopsy of the left biceps and left forearm revealed group A streptococcus within the muscle and evidence of necrosis (Figure 1). Debridement of both arms and wrists was performed. The patient subsequently developed erythema of the palms and soles followed by diffuse sloughing of the skin. Streptococcal toxic shock syndrome with necrotizing myositis was diagnosed, and the antimicrobial regimen was changed to intravenous penicillin, clindamycin, and intravenous immunoglobulin G. Repeat debridement of the arms was required.

Figure 1

Even when a soft tissue infection is suspected, it can be challenging to preoperatively localize or characterize with precision. In retrospect, the overall severity of his illness and his previously good health status perhaps favor necrotizing myositis (and fasciitis) over pyomyositis.

I may have put undue emphasis on the absence of skin findings. Although the symmetric nature of the disease is unusual, the small skin lesions may have been portals of entry, and in retrospect, they represent the tip of the iceberg. The absence of fever and leukocytosisor hypothermia and leukopeniain a young, previously healthy patient along with the bilateral and symmetric findings made me hesitant to definitely label his illness as a deep soft tissue infection early on, but the gravity of the illness, the lack of a plausible alternative explanation, and his precipitous decline all made surgical exploration imperative.

The patient's skin sloughing progressed, and he was transferred to a regional burn unit. Three additional operations were required for debridement of both upper extremities. Despite apparent control of the initial infection, the patient continued to require significant hemodynamic and ventilator support. He subsequently developed neutropenia, thrombocytopenia, and Escherichia coli urosepsis. Necrosis of the lower extremities developed, and additional surgical debridement was recommended. After extensive discussion regarding prognosis, the family decided against further surgery and withdrew life support. The patient died shortly after extubation. The family refused an autopsy.

Commentary

Expert clinicians employ a variety of approaches to solve complex clinical problems. One of the most effective strategies is pattern recognition, in which the clinician divides the case into recognizable portions and compares these to previous cases that he or she has encountered.1, 2 If patterns from the new case appear similar or identical to those of previous cases, a diagnosis can be made quickly and without the need for unnecessary testing. However, when features of the case are unusual or atypical, pattern recognition may be disrupted. For example, although the discussant suspected a soft tissue infection, a number of features (including a normal white blood cell count, minimal skin findings, and a bilateral and symmetric distribution of swelling) did not match his illness script (a mental representation of a disease) of necrotizing fasciitis.

When pattern recognition fails, other strategies are available.3 Hypothetico‐deductive reasoning is a data‐to‐diagnosis method whereby the clinician uses the presenting information to construct a list of diagnostic possibilities.4 Additional testing and gathering of information are then used to continuously revise the diagnostic possibilities until confirmatory information is obtained and the diagnosis is established. After a pattern failed to materialize, the discussant employed this analytical strategy by noting the unusual characteristics of the case and incorporating laboratory and physiological data to revise his differential diagnosis. As a result, he requested the appropriate diagnostic test: surgical exploration of the forearms.

Necrotizing soft tissue infections are characterized by fulminant tissue destruction, rapid spread along tissue planes, and local vascular thrombosis. Mixed aerobic and anaerobic infections typically occur after penetrating skin injury or following surgery in patients with diabetes mellitus or vascular disease. In contrast, monomicrobial infections with S. aureus or group A streptococcus generally occur in healthy individuals. The prevalence of necrotizing group A streptococcal infections has increased dramatically in the last 15 to 20 years.5 Over one‐third of these cases are complicated by toxic shock syndrome.57 Mortality rates for necrotizing fasciitis with toxic shock exceed 30%, and early surgical consultation is directly associated with a reduction in morbidity and mortality.5, 8

Toxic shock is an inflammatory response syndrome caused by release of exotoxins from group A streptococcus and S. aureus.9, 10 In streptococcal toxic shock, Streptococcus pyogenes exotoxin A and Streptococcus pyogenes exotoxin B are the major toxins produced.10 These toxins activate the systemic production of inflammatory cytokines such as interleukin‐1, gamma‐interferon, and tumor necrosis factor, resulting in capillary leak, systemic hypotension, tissue hypoperfusion, and organ failure. The most common initial symptom is diffuse or localized pain that is severe and abrupt in onset and often precedes or is out of proportion to other physical findings of soft tissue infection.8 Up to 20% of individuals may also develop a viral‐like syndrome with myalgias, fever, nausea, vomiting, and diarrhea.11 Erythroderma of the skin and mucous membranes can be another early finding. The rash is diffuse, erythematous, and macular, resembling a sunburn. It involves the palms and soles but can be subtle and fleeting. Erythroderma may be particularly difficult to detect in dark‐skinned individuals.12 It is also important to consider that the absence of fever, erythroderma, or leukocytosis does not necessarily rule out the possibility of serious infection in necrotizing fasciitis patients, as the development of these signs and symptoms may occur later in the disease process.

The most common portals of entry for group A streptococcus are the skin, vagina, and pharynx. Predisposing factors include varicella infection, penetrating injuries, minor cuts, burns, splinters, and surgery. Interestingly, a portal of entry cannot be identified in 45% of cases.8 These patients in particular are at risk for developing severe necrotizing myositis or fasciitis at the site of a minor injury such as a strained muscle. Hematogenous translocation from the pharynx to the site of injury is the probable mechanism13 and would provide one scenario by which our piano mover developed bilateral and symmetric disease. An alternative explanation would be direct extension of bacteria from small breaks in the skin to adjacent areas of muscle strain. Regardless of the portal of entry, as the small skin lesions demonstrate in this patient, the smallest physical finding can represent the tip of the iceberg.

Venous thromboembolism (VTE) is a common complication of serious illness, conferring increased morbidity and mortality in hospitalized medical patients. Thromboprophylaxis has been rated the number 1 patient safety intervention for hospitalized patients by the Agency for Healthcare Research and Quality, supported by evidence of effectiveness in multiple methodologically rigorous randomized trials.1 Unfortunately, many studies have shown that suitable patients do not receive thromboprophylaxis when they should. For example, in a large teaching hospital, 44 of 245 VTE events were considered to be potentially preventable, occurring because of omitted prophylaxis, inadequate duration of prophylaxis, or incorrect type of prophylaxis.2

Medical patients appear to be at particularly high risk of not receiving thromboprophylaxis. One retrospective study of 446 medical patients in 2 hospitals revealed that only 33% had appropriate prophylaxis.3 In a large retrospective study of 29 Canadian hospitals, among 1,894 patients for whom thromboprophylaxis was considered necessary, only 23% received it.4 These findings are comparable to other practice audits512 and a large international cross‐sectional study13 showing that appropriate thromboprophylaxis is administered to only 1 of every 3 hospitalized medical patients eligible for prophylaxis. Furthermore, of patients diagnosed with VTE in a large international registry, only 33% had received heparin thromboprophylaxis prior to their event.14

The objective of this study was to understand the barriers to, and facilitators of, optimal thromboprophylaxis in hospitalized medical patients using grounded theory methods. To address our poor understanding of low rates of thromboprophylaxis on hospital medical wards, we used qualitative research, which focuses on social and interpreted, rather than natural and objectified, phenomena. Qualitative research also aims to discover, describe, and understand, rather than to test and evaluate.15 Our research question was, What do healthcare clinicians, managers and hospital administrators perceive inhibits the implementation of thromboprophylaxis for medical patients, and what do they perceive would help to optimize thromboprophylaxis?

Methods

We conducted the Qualitative Thromboprophylaxis Enquiry of Compliance (QUALITEC) survey at 3 institutions from June 2006 to September 2007. One was a university‐affiliated hospital (St. Joseph's Healthcare, Hamilton, Ontario, Canada); the other 2 were community hospitals (Joseph Brant Hospital, Burlington, Ontario, and Credit Valley Hospital, Mississauga, Ontario, Canada).

Participants included: (1) bedside nurses, nurse clinicians, and nurse educators; (2) pharmacists; (3) attending physicians; (4) nurse, pharmacy, and physician managers; (5) hospital administrators; (6)medical residents; and (7) members of the hospitals' Quality Improvement Team. Participants in groups 1 to 3 were persons who had cared for medical patients in the index hospital for at least 2 years. Participants in groups 4 and 5 were persons who had worked in the index hospital for at least 2 years. The requisite minimal duration of work on the medical ward for residents was 1 month. There were no exclusion criteria for participants based on sex, religion, ethnicity, or culture. Thus, we enrolled participants based on their a priori eligibility criteria (criterion sampling) and their ability to allow us to achieve our objectives (purposive sampling).

The Research Coordinator created a list of potential interview candidates by examining the organizational structure of each hospital to identify the names of clinicians, managers, and administrators who had responsibility for medical patients. This person identified the names of potential medical ward nurses through discussion with the nurse manager of each medical unit and identified the names of potential attending physicians, residents, and pharmacists through discussion with the physician director or designate of the medical unit (snowball sampling).

After piloting testing with 2 trial participants, the Research Coordinator conducted in‐depth, open‐ended, 1‐on‐1 interviews using a flexible interview guide. This approach allows respondents to use their own words to express ideas, and affords opportunities to probe for more information through dialog. All potential participants were provided with a brief study summary. Each interview took approximately 30 to 60 minutes. All interviews were audiotaped and transcribed verbatim. The Research Coordinator wrote field notes on the main concepts featured in the interview.

To characterize the participants, we recorded their age, sex, position, number of years in their current position, number of years in their profession, and number of years working in their hospital. For clinicians, we recorded the proportion of their professional time spent in clinical practice, the proportion of their clinical time spent caring for hospitalized medical patients (on the medical wards, doing consultations, and in the emergency department), and whether they had a formal role in a previous quality improvement project.

We transcribed each interview upon completion. As is typical for qualitative research, we began data analysis during the data collection phase. Transcripts were analyzed using the coding methods described by Strauss and Corbin.16 The main content analysis of QUALITEC consisted of line‐by‐line open coding. All transcripts were coded by 2 investigators independently who created categories and themes. The first coding process of grounded theory analysis was reduction of data to identify categories (experiences and actions that were similar or related) and to define their dimensions. We iterated between data collection and analysis, revising the interview guide to refine questions and focus on unique concepts. The qualitative data management software NVivo helped with linking codes and categories to ultimately organize ideas into a main theme and several key concepts; it also helped with retrieval of specific quotes. A third investigator reviewed the transcripts independently to identify categories. Analysis continued while the study progressed until saturation of information occurred and no new categories emerged. We invited 3 participants to review an early draft report (member checking).

Research Ethics

This study was approved by the Research Ethics Board at each participating center. Written informed consent was obtained from each participant. Participation was voluntary and confidential; participants had the opportunity to withdraw from the study without any consequences. Data were deidentified upon transcription and remained anonymized, while kept in secure password protected computerized files. Participants received a $25 gift certificate as a token of appreciation. We conducted this study according to the Guiding Ethical Principles section of the Tri‐council Policy Statement of Ethical Conduct for Research Involving Humans.17

Results

Of 39 persons we approached to participate, 36 out of 39 (92.3%) agreed. Two nurse managers at 1 community hospital declined to participate. One resident was missed due to clinical responsibilities. In Table 1, we present the characteristics of the 36 participants. When asked about the level of concern about thromboprophylaxis on the medical wards, participants stated that clinicians were generally insufficiently concerned (24/36, 68.6%) or appropriately concerned (11/36, 31.4%).

Discussion

In this qualitative study, participants affirmed that depending on individual physicians for VTE prevention is insufficient. Distinct from most therapeutic interventions which are understood as the responsibility of physicians, preventive interventions such as thromboprophylaxis may be more readily embraced as the charge of members of a multidisciplinary team. While every clinician group felt compelled to help with VTE prevention, reliance on multidisciplinary care was also perceived as a barrier to effective VTE prevention because it can generate confusion about roles. Participants recommended a comprehensive, systems‐approach, including screening and risk‐stratifying all patients, preprinted orders at hospital admission that are regularly reevaluated, and audit and feedback programs. Also endorsed were patient or family‐mediated reminders, and administrative interventions such as hiring more physiotherapists for mobilization and profiling thromboprophylaxis for hospital accreditation.

Approximately 70% of participants judged that clinicians were insufficiently concerned about thromboprophylaxis. In contrast to many commonly cited reasons for underutilization of evidence‐based interventions such as lack of awareness of (or resistance to) new information, and lack of self‐efficacy of clinicians (wondering whether the benefits observed in the research setting will be realized in the practice setting),18 participants in our study did not raise these as barriers to thromboprophylaxis. As physician and pharmacist participants indicated, because the evidentiary basis for heparin thromboprophylaxis is strong and understood (in contrast to that for mechanical prophylaxis), lack of knowledge was not considered a barrier.

Our findings are consistent with a previous qualitative study in which clinicians and managers were interviewed to learn about factors that increase beta‐blocker use following myocardial infarction; these investigators found that administrative support, use of data, and quality improvement initiatives were key.19 Detailed implementation directives to clarify clinical responsibilities for many different stakeholders were also suggested in a qualitative study on the optimal use of noninvasive ventilation.20 Extending these results, participants in our study suggested several interventions aimed at different levels of the healthcare system (eg, patient, provider, and administrator) that are either enabling or reinforcing. While just 1 implementation strategy may result in improved thromboprophylaxis,21 efficient application of multiply‐redundant strategies delivered by a multidisciplinary team may be more powerful. Such an approach has driven high level performance for the American Heart Association's Get With The Guidelines Program, resulting in lower community and hospital cardiovascular and cerebrovascular mortality in Massachusetts.22

Although our aim was to obtain multidisciplinary input, we did not focus on physicians who primarily prescribe thromboprophylaxis, thereby potentially underrepresenting their views relative to their role. We did not interview all clinicians who can influence VTE risk (eg, physiotherapists) or prescribe thromboprophylaxis (eg, family physicians). During this study, patient safety emerged as a major hospital initiative,23 which could modify participants' views on the importance of VTE prevention. Although VTE prevention may be enhanced by understanding barriers to, and solutions for, optimal thromboprophylaxis from both clinicians' and managers' perspectives, the suggestions we elicited on methods to improve VTE prevention represent participants' opinions, rather than evidence, about effective strategies. Our findings are not generalizable to settings with nurse practitioners who are dedicated to VTE prevention, nor to settings with computer decision support systems that already incorporate thromboprophylaxis.

Strengths of this study include interviewing key stakeholders representing several clinician groups and managers, to obtain perspectives on both individual and systemwide influences on thromboprophylaxis in medical patients. By eliciting the views and experiences of participants in both the community and university settings, we captured multicenter perspectives on preventive health. We used triangulation of data sources (researcher and participant) and invited participants to review an early draft of the report (member checking).24 This study highlights the merits of qualitative research which can provide insights into familiar patterns and problems, and contribute to knowledge for interdisciplinary audiences. These qualitative research results help to explain quantitative research results documenting very low rates of medical thromboprophylaxis,214 raising hypotheses to test in future studies testing systemwide interventions that may improve patient safety. One prominent publication reported the impact of electronic reminders,25 while a recent systematic review of methods to increase thromboprophylaxis outlines many potentially effective approaches.21

In summary, the findings of this qualitative study challenge the notion that either individual physician efforts or multidisciplinary care are enough to lead to optimal VTE prevention. Participants believed that while well‐functioning teams at the bedside hold great promise to deliver superior care, they may also lead to unclear role definition. Since physicians commit many errors of omission regarding thromboprophylaxis, these results raise the possibility of delegated medical acts outside the scope of conventional practice for nurses and pharmacists. Leveraging the skills and knowledge of multidisciplinary teams for medical thromboprophylaxis requires not only clear accountabilities, but also excellent interprofessional communication, and institution‐wide approaches to change prescribing behavior, potentially involving patients and administrators as well as clinicians.

Venous thromboembolism (VTE) is a common complication of serious illness, conferring increased morbidity and mortality in hospitalized medical patients. Thromboprophylaxis has been rated the number 1 patient safety intervention for hospitalized patients by the Agency for Healthcare Research and Quality, supported by evidence of effectiveness in multiple methodologically rigorous randomized trials.1 Unfortunately, many studies have shown that suitable patients do not receive thromboprophylaxis when they should. For example, in a large teaching hospital, 44 of 245 VTE events were considered to be potentially preventable, occurring because of omitted prophylaxis, inadequate duration of prophylaxis, or incorrect type of prophylaxis.2

Medical patients appear to be at particularly high risk of not receiving thromboprophylaxis. One retrospective study of 446 medical patients in 2 hospitals revealed that only 33% had appropriate prophylaxis.3 In a large retrospective study of 29 Canadian hospitals, among 1,894 patients for whom thromboprophylaxis was considered necessary, only 23% received it.4 These findings are comparable to other practice audits512 and a large international cross‐sectional study13 showing that appropriate thromboprophylaxis is administered to only 1 of every 3 hospitalized medical patients eligible for prophylaxis. Furthermore, of patients diagnosed with VTE in a large international registry, only 33% had received heparin thromboprophylaxis prior to their event.14

The objective of this study was to understand the barriers to, and facilitators of, optimal thromboprophylaxis in hospitalized medical patients using grounded theory methods. To address our poor understanding of low rates of thromboprophylaxis on hospital medical wards, we used qualitative research, which focuses on social and interpreted, rather than natural and objectified, phenomena. Qualitative research also aims to discover, describe, and understand, rather than to test and evaluate.15 Our research question was, What do healthcare clinicians, managers and hospital administrators perceive inhibits the implementation of thromboprophylaxis for medical patients, and what do they perceive would help to optimize thromboprophylaxis?

Methods

We conducted the Qualitative Thromboprophylaxis Enquiry of Compliance (QUALITEC) survey at 3 institutions from June 2006 to September 2007. One was a university‐affiliated hospital (St. Joseph's Healthcare, Hamilton, Ontario, Canada); the other 2 were community hospitals (Joseph Brant Hospital, Burlington, Ontario, and Credit Valley Hospital, Mississauga, Ontario, Canada).

Participants included: (1) bedside nurses, nurse clinicians, and nurse educators; (2) pharmacists; (3) attending physicians; (4) nurse, pharmacy, and physician managers; (5) hospital administrators; (6)medical residents; and (7) members of the hospitals' Quality Improvement Team. Participants in groups 1 to 3 were persons who had cared for medical patients in the index hospital for at least 2 years. Participants in groups 4 and 5 were persons who had worked in the index hospital for at least 2 years. The requisite minimal duration of work on the medical ward for residents was 1 month. There were no exclusion criteria for participants based on sex, religion, ethnicity, or culture. Thus, we enrolled participants based on their a priori eligibility criteria (criterion sampling) and their ability to allow us to achieve our objectives (purposive sampling).

The Research Coordinator created a list of potential interview candidates by examining the organizational structure of each hospital to identify the names of clinicians, managers, and administrators who had responsibility for medical patients. This person identified the names of potential medical ward nurses through discussion with the nurse manager of each medical unit and identified the names of potential attending physicians, residents, and pharmacists through discussion with the physician director or designate of the medical unit (snowball sampling).

After piloting testing with 2 trial participants, the Research Coordinator conducted in‐depth, open‐ended, 1‐on‐1 interviews using a flexible interview guide. This approach allows respondents to use their own words to express ideas, and affords opportunities to probe for more information through dialog. All potential participants were provided with a brief study summary. Each interview took approximately 30 to 60 minutes. All interviews were audiotaped and transcribed verbatim. The Research Coordinator wrote field notes on the main concepts featured in the interview.

To characterize the participants, we recorded their age, sex, position, number of years in their current position, number of years in their profession, and number of years working in their hospital. For clinicians, we recorded the proportion of their professional time spent in clinical practice, the proportion of their clinical time spent caring for hospitalized medical patients (on the medical wards, doing consultations, and in the emergency department), and whether they had a formal role in a previous quality improvement project.

We transcribed each interview upon completion. As is typical for qualitative research, we began data analysis during the data collection phase. Transcripts were analyzed using the coding methods described by Strauss and Corbin.16 The main content analysis of QUALITEC consisted of line‐by‐line open coding. All transcripts were coded by 2 investigators independently who created categories and themes. The first coding process of grounded theory analysis was reduction of data to identify categories (experiences and actions that were similar or related) and to define their dimensions. We iterated between data collection and analysis, revising the interview guide to refine questions and focus on unique concepts. The qualitative data management software NVivo helped with linking codes and categories to ultimately organize ideas into a main theme and several key concepts; it also helped with retrieval of specific quotes. A third investigator reviewed the transcripts independently to identify categories. Analysis continued while the study progressed until saturation of information occurred and no new categories emerged. We invited 3 participants to review an early draft report (member checking).

Research Ethics

This study was approved by the Research Ethics Board at each participating center. Written informed consent was obtained from each participant. Participation was voluntary and confidential; participants had the opportunity to withdraw from the study without any consequences. Data were deidentified upon transcription and remained anonymized, while kept in secure password protected computerized files. Participants received a $25 gift certificate as a token of appreciation. We conducted this study according to the Guiding Ethical Principles section of the Tri‐council Policy Statement of Ethical Conduct for Research Involving Humans.17

Results

Of 39 persons we approached to participate, 36 out of 39 (92.3%) agreed. Two nurse managers at 1 community hospital declined to participate. One resident was missed due to clinical responsibilities. In Table 1, we present the characteristics of the 36 participants. When asked about the level of concern about thromboprophylaxis on the medical wards, participants stated that clinicians were generally insufficiently concerned (24/36, 68.6%) or appropriately concerned (11/36, 31.4%).

Discussion

In this qualitative study, participants affirmed that depending on individual physicians for VTE prevention is insufficient. Distinct from most therapeutic interventions which are understood as the responsibility of physicians, preventive interventions such as thromboprophylaxis may be more readily embraced as the charge of members of a multidisciplinary team. While every clinician group felt compelled to help with VTE prevention, reliance on multidisciplinary care was also perceived as a barrier to effective VTE prevention because it can generate confusion about roles. Participants recommended a comprehensive, systems‐approach, including screening and risk‐stratifying all patients, preprinted orders at hospital admission that are regularly reevaluated, and audit and feedback programs. Also endorsed were patient or family‐mediated reminders, and administrative interventions such as hiring more physiotherapists for mobilization and profiling thromboprophylaxis for hospital accreditation.

Approximately 70% of participants judged that clinicians were insufficiently concerned about thromboprophylaxis. In contrast to many commonly cited reasons for underutilization of evidence‐based interventions such as lack of awareness of (or resistance to) new information, and lack of self‐efficacy of clinicians (wondering whether the benefits observed in the research setting will be realized in the practice setting),18 participants in our study did not raise these as barriers to thromboprophylaxis. As physician and pharmacist participants indicated, because the evidentiary basis for heparin thromboprophylaxis is strong and understood (in contrast to that for mechanical prophylaxis), lack of knowledge was not considered a barrier.

Our findings are consistent with a previous qualitative study in which clinicians and managers were interviewed to learn about factors that increase beta‐blocker use following myocardial infarction; these investigators found that administrative support, use of data, and quality improvement initiatives were key.19 Detailed implementation directives to clarify clinical responsibilities for many different stakeholders were also suggested in a qualitative study on the optimal use of noninvasive ventilation.20 Extending these results, participants in our study suggested several interventions aimed at different levels of the healthcare system (eg, patient, provider, and administrator) that are either enabling or reinforcing. While just 1 implementation strategy may result in improved thromboprophylaxis,21 efficient application of multiply‐redundant strategies delivered by a multidisciplinary team may be more powerful. Such an approach has driven high level performance for the American Heart Association's Get With The Guidelines Program, resulting in lower community and hospital cardiovascular and cerebrovascular mortality in Massachusetts.22

Although our aim was to obtain multidisciplinary input, we did not focus on physicians who primarily prescribe thromboprophylaxis, thereby potentially underrepresenting their views relative to their role. We did not interview all clinicians who can influence VTE risk (eg, physiotherapists) or prescribe thromboprophylaxis (eg, family physicians). During this study, patient safety emerged as a major hospital initiative,23 which could modify participants' views on the importance of VTE prevention. Although VTE prevention may be enhanced by understanding barriers to, and solutions for, optimal thromboprophylaxis from both clinicians' and managers' perspectives, the suggestions we elicited on methods to improve VTE prevention represent participants' opinions, rather than evidence, about effective strategies. Our findings are not generalizable to settings with nurse practitioners who are dedicated to VTE prevention, nor to settings with computer decision support systems that already incorporate thromboprophylaxis.

Strengths of this study include interviewing key stakeholders representing several clinician groups and managers, to obtain perspectives on both individual and systemwide influences on thromboprophylaxis in medical patients. By eliciting the views and experiences of participants in both the community and university settings, we captured multicenter perspectives on preventive health. We used triangulation of data sources (researcher and participant) and invited participants to review an early draft of the report (member checking).24 This study highlights the merits of qualitative research which can provide insights into familiar patterns and problems, and contribute to knowledge for interdisciplinary audiences. These qualitative research results help to explain quantitative research results documenting very low rates of medical thromboprophylaxis,214 raising hypotheses to test in future studies testing systemwide interventions that may improve patient safety. One prominent publication reported the impact of electronic reminders,25 while a recent systematic review of methods to increase thromboprophylaxis outlines many potentially effective approaches.21

In summary, the findings of this qualitative study challenge the notion that either individual physician efforts or multidisciplinary care are enough to lead to optimal VTE prevention. Participants believed that while well‐functioning teams at the bedside hold great promise to deliver superior care, they may also lead to unclear role definition. Since physicians commit many errors of omission regarding thromboprophylaxis, these results raise the possibility of delegated medical acts outside the scope of conventional practice for nurses and pharmacists. Leveraging the skills and knowledge of multidisciplinary teams for medical thromboprophylaxis requires not only clear accountabilities, but also excellent interprofessional communication, and institution‐wide approaches to change prescribing behavior, potentially involving patients and administrators as well as clinicians.

Venous thromboembolism (VTE) is a common complication of serious illness, conferring increased morbidity and mortality in hospitalized medical patients. Thromboprophylaxis has been rated the number 1 patient safety intervention for hospitalized patients by the Agency for Healthcare Research and Quality, supported by evidence of effectiveness in multiple methodologically rigorous randomized trials.1 Unfortunately, many studies have shown that suitable patients do not receive thromboprophylaxis when they should. For example, in a large teaching hospital, 44 of 245 VTE events were considered to be potentially preventable, occurring because of omitted prophylaxis, inadequate duration of prophylaxis, or incorrect type of prophylaxis.2

Medical patients appear to be at particularly high risk of not receiving thromboprophylaxis. One retrospective study of 446 medical patients in 2 hospitals revealed that only 33% had appropriate prophylaxis.3 In a large retrospective study of 29 Canadian hospitals, among 1,894 patients for whom thromboprophylaxis was considered necessary, only 23% received it.4 These findings are comparable to other practice audits512 and a large international cross‐sectional study13 showing that appropriate thromboprophylaxis is administered to only 1 of every 3 hospitalized medical patients eligible for prophylaxis. Furthermore, of patients diagnosed with VTE in a large international registry, only 33% had received heparin thromboprophylaxis prior to their event.14

The objective of this study was to understand the barriers to, and facilitators of, optimal thromboprophylaxis in hospitalized medical patients using grounded theory methods. To address our poor understanding of low rates of thromboprophylaxis on hospital medical wards, we used qualitative research, which focuses on social and interpreted, rather than natural and objectified, phenomena. Qualitative research also aims to discover, describe, and understand, rather than to test and evaluate.15 Our research question was, What do healthcare clinicians, managers and hospital administrators perceive inhibits the implementation of thromboprophylaxis for medical patients, and what do they perceive would help to optimize thromboprophylaxis?

Methods

We conducted the Qualitative Thromboprophylaxis Enquiry of Compliance (QUALITEC) survey at 3 institutions from June 2006 to September 2007. One was a university‐affiliated hospital (St. Joseph's Healthcare, Hamilton, Ontario, Canada); the other 2 were community hospitals (Joseph Brant Hospital, Burlington, Ontario, and Credit Valley Hospital, Mississauga, Ontario, Canada).

Participants included: (1) bedside nurses, nurse clinicians, and nurse educators; (2) pharmacists; (3) attending physicians; (4) nurse, pharmacy, and physician managers; (5) hospital administrators; (6)medical residents; and (7) members of the hospitals' Quality Improvement Team. Participants in groups 1 to 3 were persons who had cared for medical patients in the index hospital for at least 2 years. Participants in groups 4 and 5 were persons who had worked in the index hospital for at least 2 years. The requisite minimal duration of work on the medical ward for residents was 1 month. There were no exclusion criteria for participants based on sex, religion, ethnicity, or culture. Thus, we enrolled participants based on their a priori eligibility criteria (criterion sampling) and their ability to allow us to achieve our objectives (purposive sampling).

The Research Coordinator created a list of potential interview candidates by examining the organizational structure of each hospital to identify the names of clinicians, managers, and administrators who had responsibility for medical patients. This person identified the names of potential medical ward nurses through discussion with the nurse manager of each medical unit and identified the names of potential attending physicians, residents, and pharmacists through discussion with the physician director or designate of the medical unit (snowball sampling).

After piloting testing with 2 trial participants, the Research Coordinator conducted in‐depth, open‐ended, 1‐on‐1 interviews using a flexible interview guide. This approach allows respondents to use their own words to express ideas, and affords opportunities to probe for more information through dialog. All potential participants were provided with a brief study summary. Each interview took approximately 30 to 60 minutes. All interviews were audiotaped and transcribed verbatim. The Research Coordinator wrote field notes on the main concepts featured in the interview.

To characterize the participants, we recorded their age, sex, position, number of years in their current position, number of years in their profession, and number of years working in their hospital. For clinicians, we recorded the proportion of their professional time spent in clinical practice, the proportion of their clinical time spent caring for hospitalized medical patients (on the medical wards, doing consultations, and in the emergency department), and whether they had a formal role in a previous quality improvement project.

We transcribed each interview upon completion. As is typical for qualitative research, we began data analysis during the data collection phase. Transcripts were analyzed using the coding methods described by Strauss and Corbin.16 The main content analysis of QUALITEC consisted of line‐by‐line open coding. All transcripts were coded by 2 investigators independently who created categories and themes. The first coding process of grounded theory analysis was reduction of data to identify categories (experiences and actions that were similar or related) and to define their dimensions. We iterated between data collection and analysis, revising the interview guide to refine questions and focus on unique concepts. The qualitative data management software NVivo helped with linking codes and categories to ultimately organize ideas into a main theme and several key concepts; it also helped with retrieval of specific quotes. A third investigator reviewed the transcripts independently to identify categories. Analysis continued while the study progressed until saturation of information occurred and no new categories emerged. We invited 3 participants to review an early draft report (member checking).

Research Ethics

This study was approved by the Research Ethics Board at each participating center. Written informed consent was obtained from each participant. Participation was voluntary and confidential; participants had the opportunity to withdraw from the study without any consequences. Data were deidentified upon transcription and remained anonymized, while kept in secure password protected computerized files. Participants received a $25 gift certificate as a token of appreciation. We conducted this study according to the Guiding Ethical Principles section of the Tri‐council Policy Statement of Ethical Conduct for Research Involving Humans.17

Results

Of 39 persons we approached to participate, 36 out of 39 (92.3%) agreed. Two nurse managers at 1 community hospital declined to participate. One resident was missed due to clinical responsibilities. In Table 1, we present the characteristics of the 36 participants. When asked about the level of concern about thromboprophylaxis on the medical wards, participants stated that clinicians were generally insufficiently concerned (24/36, 68.6%) or appropriately concerned (11/36, 31.4%).

Discussion

In this qualitative study, participants affirmed that depending on individual physicians for VTE prevention is insufficient. Distinct from most therapeutic interventions which are understood as the responsibility of physicians, preventive interventions such as thromboprophylaxis may be more readily embraced as the charge of members of a multidisciplinary team. While every clinician group felt compelled to help with VTE prevention, reliance on multidisciplinary care was also perceived as a barrier to effective VTE prevention because it can generate confusion about roles. Participants recommended a comprehensive, systems‐approach, including screening and risk‐stratifying all patients, preprinted orders at hospital admission that are regularly reevaluated, and audit and feedback programs. Also endorsed were patient or family‐mediated reminders, and administrative interventions such as hiring more physiotherapists for mobilization and profiling thromboprophylaxis for hospital accreditation.

Approximately 70% of participants judged that clinicians were insufficiently concerned about thromboprophylaxis. In contrast to many commonly cited reasons for underutilization of evidence‐based interventions such as lack of awareness of (or resistance to) new information, and lack of self‐efficacy of clinicians (wondering whether the benefits observed in the research setting will be realized in the practice setting),18 participants in our study did not raise these as barriers to thromboprophylaxis. As physician and pharmacist participants indicated, because the evidentiary basis for heparin thromboprophylaxis is strong and understood (in contrast to that for mechanical prophylaxis), lack of knowledge was not considered a barrier.

Our findings are consistent with a previous qualitative study in which clinicians and managers were interviewed to learn about factors that increase beta‐blocker use following myocardial infarction; these investigators found that administrative support, use of data, and quality improvement initiatives were key.19 Detailed implementation directives to clarify clinical responsibilities for many different stakeholders were also suggested in a qualitative study on the optimal use of noninvasive ventilation.20 Extending these results, participants in our study suggested several interventions aimed at different levels of the healthcare system (eg, patient, provider, and administrator) that are either enabling or reinforcing. While just 1 implementation strategy may result in improved thromboprophylaxis,21 efficient application of multiply‐redundant strategies delivered by a multidisciplinary team may be more powerful. Such an approach has driven high level performance for the American Heart Association's Get With The Guidelines Program, resulting in lower community and hospital cardiovascular and cerebrovascular mortality in Massachusetts.22

Although our aim was to obtain multidisciplinary input, we did not focus on physicians who primarily prescribe thromboprophylaxis, thereby potentially underrepresenting their views relative to their role. We did not interview all clinicians who can influence VTE risk (eg, physiotherapists) or prescribe thromboprophylaxis (eg, family physicians). During this study, patient safety emerged as a major hospital initiative,23 which could modify participants' views on the importance of VTE prevention. Although VTE prevention may be enhanced by understanding barriers to, and solutions for, optimal thromboprophylaxis from both clinicians' and managers' perspectives, the suggestions we elicited on methods to improve VTE prevention represent participants' opinions, rather than evidence, about effective strategies. Our findings are not generalizable to settings with nurse practitioners who are dedicated to VTE prevention, nor to settings with computer decision support systems that already incorporate thromboprophylaxis.

Strengths of this study include interviewing key stakeholders representing several clinician groups and managers, to obtain perspectives on both individual and systemwide influences on thromboprophylaxis in medical patients. By eliciting the views and experiences of participants in both the community and university settings, we captured multicenter perspectives on preventive health. We used triangulation of data sources (researcher and participant) and invited participants to review an early draft of the report (member checking).24 This study highlights the merits of qualitative research which can provide insights into familiar patterns and problems, and contribute to knowledge for interdisciplinary audiences. These qualitative research results help to explain quantitative research results documenting very low rates of medical thromboprophylaxis,214 raising hypotheses to test in future studies testing systemwide interventions that may improve patient safety. One prominent publication reported the impact of electronic reminders,25 while a recent systematic review of methods to increase thromboprophylaxis outlines many potentially effective approaches.21

In summary, the findings of this qualitative study challenge the notion that either individual physician efforts or multidisciplinary care are enough to lead to optimal VTE prevention. Participants believed that while well‐functioning teams at the bedside hold great promise to deliver superior care, they may also lead to unclear role definition. Since physicians commit many errors of omission regarding thromboprophylaxis, these results raise the possibility of delegated medical acts outside the scope of conventional practice for nurses and pharmacists. Leveraging the skills and knowledge of multidisciplinary teams for medical thromboprophylaxis requires not only clear accountabilities, but also excellent interprofessional communication, and institution‐wide approaches to change prescribing behavior, potentially involving patients and administrators as well as clinicians.

A 22‐year‐old man was referred from an outside hospital with 2 weeks of generalized weakness and difficulty ambulating. He also reported a 6‐month history of abdominal pain, right‐sided lumbar back pain, malaise, night sweats, cough, and weight loss of 30 kg. The patient was born in Mexico but had lived in the US for 10 years, working as a construction worker in northern California. He had had no prior medical care. On examination, the patient had a temperature of 40C, a heart rate of 109, and a respiratory rate of 20. His oxygen saturation was 97% on room air, and lung fields were clear bilaterally. Strength was slightly decreased in the right leg, with tenderness in the posterolateral aspect of the right thigh and thoracic lumbar spine.

Radiograph of the chest showed bilateral upper lobe reticulonodular infiltrates, cavitation, and pleural thickening (Figure 1). Computed tomography (CT) showed innumerable pulmonary nodules throughout the mid and lower lungs, a loss of disc space between L4 and L5, and a large multiloculated abscess of the right iliacus and psoas muscles that extended into the right thigh (Figure 2). Longitudinal relaxation time (T₁)‐weighted magnetic resonance imaging (MRI) of the lumbar spine revealed vertebral destruction of L4 and L5 (Figure 3).

Figure 1

Figure 2

Figure 3

The patient underwent vertebrectomy of L4 and L5, anterior/posterior fixation and fusion from L4 to S1, and drainage of a large right‐sided psoas abscess. Acid‐fast bacilli were seen in abscess fluid and vertebral bone. Subsequently, the intraoperative cultures and 6 sequential sputum cultures all grew Mycobacterium tuberculosis. The patient was begun on rifampin, isoniazid, pyrazinamide, and ethambutol with a good clinical response, and subsequently transferred back to his referring hospital for continued medical care and rehabilitation.

Extrapulmonary manifestations of tuberculosis should be suspected in patients from a tuberculosis‐endemic country of origin. Bone and joint tuberculosis account for up to 35% of cases of extrapulmonary tuberculosis. Spinal tuberculosis (Pott's disease) most commonly involves the anteroinferior aspect of vertebral bodies in the thoracic spine. Tuberculosis is a disease with diverse manifestations and can elude even the most astute physician if it is not considered as a diagnosis.

A 22‐year‐old man was referred from an outside hospital with 2 weeks of generalized weakness and difficulty ambulating. He also reported a 6‐month history of abdominal pain, right‐sided lumbar back pain, malaise, night sweats, cough, and weight loss of 30 kg. The patient was born in Mexico but had lived in the US for 10 years, working as a construction worker in northern California. He had had no prior medical care. On examination, the patient had a temperature of 40C, a heart rate of 109, and a respiratory rate of 20. His oxygen saturation was 97% on room air, and lung fields were clear bilaterally. Strength was slightly decreased in the right leg, with tenderness in the posterolateral aspect of the right thigh and thoracic lumbar spine.

Radiograph of the chest showed bilateral upper lobe reticulonodular infiltrates, cavitation, and pleural thickening (Figure 1). Computed tomography (CT) showed innumerable pulmonary nodules throughout the mid and lower lungs, a loss of disc space between L4 and L5, and a large multiloculated abscess of the right iliacus and psoas muscles that extended into the right thigh (Figure 2). Longitudinal relaxation time (T₁)‐weighted magnetic resonance imaging (MRI) of the lumbar spine revealed vertebral destruction of L4 and L5 (Figure 3).

Figure 1

Figure 2

Figure 3

The patient underwent vertebrectomy of L4 and L5, anterior/posterior fixation and fusion from L4 to S1, and drainage of a large right‐sided psoas abscess. Acid‐fast bacilli were seen in abscess fluid and vertebral bone. Subsequently, the intraoperative cultures and 6 sequential sputum cultures all grew Mycobacterium tuberculosis. The patient was begun on rifampin, isoniazid, pyrazinamide, and ethambutol with a good clinical response, and subsequently transferred back to his referring hospital for continued medical care and rehabilitation.

Extrapulmonary manifestations of tuberculosis should be suspected in patients from a tuberculosis‐endemic country of origin. Bone and joint tuberculosis account for up to 35% of cases of extrapulmonary tuberculosis. Spinal tuberculosis (Pott's disease) most commonly involves the anteroinferior aspect of vertebral bodies in the thoracic spine. Tuberculosis is a disease with diverse manifestations and can elude even the most astute physician if it is not considered as a diagnosis.

A 22‐year‐old man was referred from an outside hospital with 2 weeks of generalized weakness and difficulty ambulating. He also reported a 6‐month history of abdominal pain, right‐sided lumbar back pain, malaise, night sweats, cough, and weight loss of 30 kg. The patient was born in Mexico but had lived in the US for 10 years, working as a construction worker in northern California. He had had no prior medical care. On examination, the patient had a temperature of 40C, a heart rate of 109, and a respiratory rate of 20. His oxygen saturation was 97% on room air, and lung fields were clear bilaterally. Strength was slightly decreased in the right leg, with tenderness in the posterolateral aspect of the right thigh and thoracic lumbar spine.

Radiograph of the chest showed bilateral upper lobe reticulonodular infiltrates, cavitation, and pleural thickening (Figure 1). Computed tomography (CT) showed innumerable pulmonary nodules throughout the mid and lower lungs, a loss of disc space between L4 and L5, and a large multiloculated abscess of the right iliacus and psoas muscles that extended into the right thigh (Figure 2). Longitudinal relaxation time (T₁)‐weighted magnetic resonance imaging (MRI) of the lumbar spine revealed vertebral destruction of L4 and L5 (Figure 3).

Figure 1

Figure 2

Figure 3

The patient underwent vertebrectomy of L4 and L5, anterior/posterior fixation and fusion from L4 to S1, and drainage of a large right‐sided psoas abscess. Acid‐fast bacilli were seen in abscess fluid and vertebral bone. Subsequently, the intraoperative cultures and 6 sequential sputum cultures all grew Mycobacterium tuberculosis. The patient was begun on rifampin, isoniazid, pyrazinamide, and ethambutol with a good clinical response, and subsequently transferred back to his referring hospital for continued medical care and rehabilitation.

Extrapulmonary manifestations of tuberculosis should be suspected in patients from a tuberculosis‐endemic country of origin. Bone and joint tuberculosis account for up to 35% of cases of extrapulmonary tuberculosis. Spinal tuberculosis (Pott's disease) most commonly involves the anteroinferior aspect of vertebral bodies in the thoracic spine. Tuberculosis is a disease with diverse manifestations and can elude even the most astute physician if it is not considered as a diagnosis.

Deep venous thrombosis (DVT) and pulmonary embolism (PE), collectively referred to as venous thromboembolism (VTE), are common events in hospitalized patients and result in significant morbidity and mortality. Often silent and frequently unexpected, VTE is preventable. Accordingly, the American College of Chest Physicians recommends that pharmacologic prophylaxis be given to acutely ill medical patients admitted to the hospital with congestive heart failure or severe respiratory disease, or to patients who are confined to bed who have additional risk factors, such as cancer or previous VTE.1 Three recent meta‐analyses24 demonstrated significant reductions in VTE in general medicine patients with pharmacologic prophylaxis. Recently the National Quality Forum advocated that hospitals evaluate each patient upon admission and regularly thereafter, for the risk of developing DVT/VTE and utilize clinically appropriate methods to prevent DVT/VTE.5

Despite recommendations for prophylaxis, multiple studies demonstrate utilization in <50% of at‐risk general medical patients.68 Physicians' lack of awareness may partially explain this underutilization, but other likely factors include physicians' questions about the clinical importance of the outcome (eg, some studies have shown reductions primarily in asymptomatic distal DVT), doubt regarding the best form of prophylaxis (ie, unfractionated heparin [UFH] vs. low molecular weight heparin [LMWH]), uncertainty regarding optimal dosing regimens, and comparable uncertainty regarding which patients have sufficiently high risk for VTE to outweigh the risks of anticoagulation.

We undertook the current meta‐analysis to address questions about thromboembolism prevention in general medicine patients. Does pharmacologic prophylaxis prevent clinically relevant events? Is LMWH or UFH preferable in terms of either efficacy or safety?

MATERIALS AND METHODS

RESULTS

Study Identification and Selection

The computerized literature search resulted in 5284 articles. Three additional citations were found by review of bibliographies. No additional trials were identified from reviews of abstracts from national meetings. Representatives from the 3 pharmaceutical companies reported no knowledge of additional published or unpublished data. Of the 5287 studies identified by the search, 14 studies met all eligibility criteria (Figure 1).

Figure 1

UFH or LMWH vs. Control

DVT

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of all DVT (RR = 0.55; 95% CI: 0.36‐0.83) (Figure 2A). When stratified by methodological quality, 5 trials16, 2225 with Jadad scores of 3 or higher showed an RR reduction of 0.53 (95% CI: 0.38‐0.72) in reducing all DVT. All of the higher‐quality trials compared LMWH to placebo. Across 4 trials that reported data for symptomatic DVT there was a nonsignificant reduction in RR compared with placebo (RR = 0.73; 95% CI: 0.45‐1.16) (Figure 2B). Only 2 trials24, 25 (both LMWH trials) reported results for proximal DVT and demonstrated significant benefit of prophylaxis with a pooled RR of 0.46 (95% CI: 0.31‐0.69) (Figure 2C).

Figure 2

PE

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of PE (RR = 0.70; 95% CI: 0.53‐0.93) (Figure 3A). The 5 trials16, 2225 with Jadad scores of 3 or greater showed a similar relative risk reduction, but the result was no longer statistically significant (RR = 0.56; 95% CI: 0.31‐1.02). Two of the trials16, 21 relied solely on the results of autopsy to diagnose PE, which may have given rise to chance differences in detection due to generally low autopsy rates. Eliminating these 2 studies from the analysis resulted in loss of statistical significance for the reduction in risk for PE (RR = 0.48; 95% CI: 0.20‐1.15).

Figure 3

Death

Seven trials16, 2025 comparing either UFH or LMWH to control examined the impact of pharmacologic prophylaxis on death and found no significant difference between treated and untreated patients across all trials (RR = 0.92; 95% CI: 0.82‐1.03) and those limited to studies with Jadad scores of 3 or higher (RR = 0.97; 95% CI: 0.80‐1.17).

Complications

We evaluated adverse events of heparin products used for prophylaxis and whether there were differences between UFH and LMWH. Reporting of complications was not uniform from study to study, making pooling more difficult. However, we were able to abstract data on any bleeding, major bleeding, and thrombocytopenia from several studies. In 5 studies15, 16, 2325 of either UFH or LMWH vs. control, a significantly increased risk of any bleeding (RR = 1.54; 95% CI: 1.15‐2.06) (Figure 4A) was found. When only major bleeding was evaluated, no statistically significant difference was noted (RR = 1.20; 95% CI: 0.55‐2.58) (Figure 4B). In 4 trials16, 22, 24, 25 the occurrence of thrombocytopenia was not significantly different when comparing UFH or LMWH to control (RR = 0.92; 95% CI: 0.46‐1.86).

Figure 4

When LMWH was compared to UFH in 4 trials,14, 17, 18, 27 a nonsignificant trend toward a decrease in any bleeding was found in the LMWH group (RR = 0.72; 95% CI: 0.44‐1.16) (Figure 5A). A similar trend was seen favoring LMWH in rates of major bleeding (RR = 0.57; 95% CI: 0.25‐1.32) (Figure 5B). Neither trend was statistically significant. Three trials comparing LMWH to UFH reported on thrombocytopenia17, 18, 27 with no significant difference noted (RR = 0.52; 95% CI: 0.06‐4.18).

Figure 5

DISCUSSION

When compared to control, LMWH or UFH decreased the risk of all DVT by 45% (RR = 0.55; 95% CI: 0.36‐0.83) and proximal DVT by 54% (RR = 0.46; 95% CI: 0.31‐0.69). PE was also decreased by 30% (RR = 0.70; 95% CI: 0.53‐0.93). Of note, when prophylaxis was compared with placebo all of the high‐quality studies showing a benefit were done using LMWH. The benefits of prophylaxis occurred at the cost of a 54% increased overall risk of bleeding (RR = 1.54; 95% CI 1.15‐2.06). However, the risk of major bleeding was not significantly increased. We did not find a mortality benefit to pharmacologic thromboembolism prophylaxis.

When comparing UFH to LMWH, we noted no difference in all DVT, symptomatic DVT, proximal DVT, PE, or death. While there was a trend toward less bleeding with LMWH, this was not statistically significant.

Taken in aggregate, our findings are in agreement with previous published meta‐analyses reporting net benefit for thromboembolism prophylaxis in medical patients.24, 22, 28, 29 Our meta‐analysis has several methodological strengths over the prior studies, including a comprehensive search of both the published and unpublished literature and assessment of the relationship between methodological quality of included trials and reported benefit. In contrast to previous reviews, our analysis highlights several limitations of the current evidence.

First, many of the studies are older, with predicted lengths of stay of greater than 1 week. The 8‐13‐day range of treatment duration we found in this study is longer than the average length of stay in today's hospitals. Second, there is variability in the diagnostic tests used to diagnose DVT, as well as variation in the definition of DVT among studies. Studies using fibrinogen uptake scanning reported rates of DVT as high as 26%15 while studies using venography reported DVT rates of almost 15% in the placebo arm.24 These rates are higher than most physicians' routine practice. One reason for this discrepancy is most studies did not distinguish below‐the‐knee DVT from more clinically relevant above‐the‐knee DVT. Systematic reviews of medical and surgical patients have found rates of proximal propagation from 0% to 29% in untreated patients.30, 31 Though controversial, below‐the‐knee DVT is believed less morbid than proximal DVT or symptomatic DVT. We addressed this by focusing specifically on clinically relevant endpoints of proximal and symptomatic DVT. When we restricted our analysis to proximal DVT we found a 54% RR reduction in 2 pooled trials of LMWH compared to placebo. In pooled analyses symptomatic DVT was not affected by prophylaxis. When compared head‐to‐head there were no differences between LMWH and UFH for proximal DVT or symptomatic DVT.

When considering PE, the utilization of autopsy as the sole diagnostic method in 2 large trials16, 21 is particularly problematic. In the trial by Garlund,21 the mortality rate was 5.4%, with an autopsy rate of 60.1%. Similarly, in the trial by Mahe et al.,16 the mortality rate was 10%, with an autopsy rate of 49%. Given the low absolute number of deaths and substantial proportion of decedents without autopsy, the potential for chance to produce an imbalance in detection of PE is high in these studies. When we excluded these 2 trials, we found that PE was no longer reduced to a statistically significant degree by prophylaxis. Loss of significance for PE in 2 sensitivity analyses (when excluding studies of lower quality, or using autopsy as a sole diagnostic study) is problematic and calls into question the true benefit of prophylaxis for prevention of PE.

Another limitation of the current literature centers on the variability of dosing used. We pooled trials of UFH whether given BID or TID. Given the small number of trials we did not do sensitivity analyses by dosage. A recent meta‐analysis3 found both doses are efficacious, while a recent review article32 suggested superiority of TID dosing. We believe the available literature does not clearly address this issue. Regarding comparisons of LMWH to UFH, dosing variability was also noted. The trial by Bergmann and Neuhart27 used enoxaparin 20 mg per day and found similar efficacy to UFH BID, while the Samama et al.24 trial found enoxaparin 20 mg per day no more efficacious than placebo. While the literature does not clearly define a best dose, we believe enoxaparin doses lower than 40 mg daily do not reflect the standard of care.

An additional limitation of the literature is publication bias. We assessed the possibility of publication bias by a variety of means. We did find statistical evidence of publication bias for the outcome of PE when prophylaxis was compared to control. Importantly, two meta‐analyses2, 4 on thromboembolism prophylaxis for general medicine patients suggested publication bias is present and our finding supports this conclusion. While no test for publication bias is foolproof, the best protection against publication bias, which we pursued in our study, consists of a thorough search for unpublished studies, including a search of conference proceedings, contact with experts in the field, and manufacturers of LMWH.

A final limitation of the current literature centers on risk assessment. All of the trials in this meta‐analysis included patients with an elevated level of risk. Unfortunately, risk was not clearly defined in many studies, and there was no minimum level of risk between trials. While immobility, age, and length of stay were reported for most studies, other risk factors such as personal history of thromboembolism and malignancy were not uniformly reported. Based on our analysis we are not confident our results can be extrapolated to all general medicine patients.

In conclusion, we found good evidence that pharmacologic prophylaxis significantly decreases the risk of all DVT and proximal DVT in at‐risk general medical patients. However, only LMWH was shown to prevent proximal DVT. We found inconclusive evidence that prophylaxis prevents PE. When compared directly we did not find clear superiority between UFH and LMWH, though several limitations of the current literature hamper decision‐making. Given the lower cost, it may seem justified to use UFH. However, there are other practical issues, such as the fact that LMWH is given once daily, and so potentially preferred by patients and more efficient for nurses. All of these results pertain to patients with elevated risk. While we did not find significant safety concerns with prophylaxis we do not know if these results can be extrapolated to lower‐risk patients. We believe that recommending widespread prophylaxis of all general medicine patients requires additional evidence about appropriate patient selection.

Deep venous thrombosis (DVT) and pulmonary embolism (PE), collectively referred to as venous thromboembolism (VTE), are common events in hospitalized patients and result in significant morbidity and mortality. Often silent and frequently unexpected, VTE is preventable. Accordingly, the American College of Chest Physicians recommends that pharmacologic prophylaxis be given to acutely ill medical patients admitted to the hospital with congestive heart failure or severe respiratory disease, or to patients who are confined to bed who have additional risk factors, such as cancer or previous VTE.1 Three recent meta‐analyses24 demonstrated significant reductions in VTE in general medicine patients with pharmacologic prophylaxis. Recently the National Quality Forum advocated that hospitals evaluate each patient upon admission and regularly thereafter, for the risk of developing DVT/VTE and utilize clinically appropriate methods to prevent DVT/VTE.5

Despite recommendations for prophylaxis, multiple studies demonstrate utilization in <50% of at‐risk general medical patients.68 Physicians' lack of awareness may partially explain this underutilization, but other likely factors include physicians' questions about the clinical importance of the outcome (eg, some studies have shown reductions primarily in asymptomatic distal DVT), doubt regarding the best form of prophylaxis (ie, unfractionated heparin [UFH] vs. low molecular weight heparin [LMWH]), uncertainty regarding optimal dosing regimens, and comparable uncertainty regarding which patients have sufficiently high risk for VTE to outweigh the risks of anticoagulation.

We undertook the current meta‐analysis to address questions about thromboembolism prevention in general medicine patients. Does pharmacologic prophylaxis prevent clinically relevant events? Is LMWH or UFH preferable in terms of either efficacy or safety?

MATERIALS AND METHODS

RESULTS

Study Identification and Selection

The computerized literature search resulted in 5284 articles. Three additional citations were found by review of bibliographies. No additional trials were identified from reviews of abstracts from national meetings. Representatives from the 3 pharmaceutical companies reported no knowledge of additional published or unpublished data. Of the 5287 studies identified by the search, 14 studies met all eligibility criteria (Figure 1).

Figure 1

UFH or LMWH vs. Control

DVT

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of all DVT (RR = 0.55; 95% CI: 0.36‐0.83) (Figure 2A). When stratified by methodological quality, 5 trials16, 2225 with Jadad scores of 3 or higher showed an RR reduction of 0.53 (95% CI: 0.38‐0.72) in reducing all DVT. All of the higher‐quality trials compared LMWH to placebo. Across 4 trials that reported data for symptomatic DVT there was a nonsignificant reduction in RR compared with placebo (RR = 0.73; 95% CI: 0.45‐1.16) (Figure 2B). Only 2 trials24, 25 (both LMWH trials) reported results for proximal DVT and demonstrated significant benefit of prophylaxis with a pooled RR of 0.46 (95% CI: 0.31‐0.69) (Figure 2C).

Figure 2

PE

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of PE (RR = 0.70; 95% CI: 0.53‐0.93) (Figure 3A). The 5 trials16, 2225 with Jadad scores of 3 or greater showed a similar relative risk reduction, but the result was no longer statistically significant (RR = 0.56; 95% CI: 0.31‐1.02). Two of the trials16, 21 relied solely on the results of autopsy to diagnose PE, which may have given rise to chance differences in detection due to generally low autopsy rates. Eliminating these 2 studies from the analysis resulted in loss of statistical significance for the reduction in risk for PE (RR = 0.48; 95% CI: 0.20‐1.15).

Figure 3

Death

Seven trials16, 2025 comparing either UFH or LMWH to control examined the impact of pharmacologic prophylaxis on death and found no significant difference between treated and untreated patients across all trials (RR = 0.92; 95% CI: 0.82‐1.03) and those limited to studies with Jadad scores of 3 or higher (RR = 0.97; 95% CI: 0.80‐1.17).

Complications

We evaluated adverse events of heparin products used for prophylaxis and whether there were differences between UFH and LMWH. Reporting of complications was not uniform from study to study, making pooling more difficult. However, we were able to abstract data on any bleeding, major bleeding, and thrombocytopenia from several studies. In 5 studies15, 16, 2325 of either UFH or LMWH vs. control, a significantly increased risk of any bleeding (RR = 1.54; 95% CI: 1.15‐2.06) (Figure 4A) was found. When only major bleeding was evaluated, no statistically significant difference was noted (RR = 1.20; 95% CI: 0.55‐2.58) (Figure 4B). In 4 trials16, 22, 24, 25 the occurrence of thrombocytopenia was not significantly different when comparing UFH or LMWH to control (RR = 0.92; 95% CI: 0.46‐1.86).

Figure 4

When LMWH was compared to UFH in 4 trials,14, 17, 18, 27 a nonsignificant trend toward a decrease in any bleeding was found in the LMWH group (RR = 0.72; 95% CI: 0.44‐1.16) (Figure 5A). A similar trend was seen favoring LMWH in rates of major bleeding (RR = 0.57; 95% CI: 0.25‐1.32) (Figure 5B). Neither trend was statistically significant. Three trials comparing LMWH to UFH reported on thrombocytopenia17, 18, 27 with no significant difference noted (RR = 0.52; 95% CI: 0.06‐4.18).

Figure 5

DISCUSSION

When compared to control, LMWH or UFH decreased the risk of all DVT by 45% (RR = 0.55; 95% CI: 0.36‐0.83) and proximal DVT by 54% (RR = 0.46; 95% CI: 0.31‐0.69). PE was also decreased by 30% (RR = 0.70; 95% CI: 0.53‐0.93). Of note, when prophylaxis was compared with placebo all of the high‐quality studies showing a benefit were done using LMWH. The benefits of prophylaxis occurred at the cost of a 54% increased overall risk of bleeding (RR = 1.54; 95% CI 1.15‐2.06). However, the risk of major bleeding was not significantly increased. We did not find a mortality benefit to pharmacologic thromboembolism prophylaxis.

When comparing UFH to LMWH, we noted no difference in all DVT, symptomatic DVT, proximal DVT, PE, or death. While there was a trend toward less bleeding with LMWH, this was not statistically significant.

Taken in aggregate, our findings are in agreement with previous published meta‐analyses reporting net benefit for thromboembolism prophylaxis in medical patients.24, 22, 28, 29 Our meta‐analysis has several methodological strengths over the prior studies, including a comprehensive search of both the published and unpublished literature and assessment of the relationship between methodological quality of included trials and reported benefit. In contrast to previous reviews, our analysis highlights several limitations of the current evidence.

First, many of the studies are older, with predicted lengths of stay of greater than 1 week. The 8‐13‐day range of treatment duration we found in this study is longer than the average length of stay in today's hospitals. Second, there is variability in the diagnostic tests used to diagnose DVT, as well as variation in the definition of DVT among studies. Studies using fibrinogen uptake scanning reported rates of DVT as high as 26%15 while studies using venography reported DVT rates of almost 15% in the placebo arm.24 These rates are higher than most physicians' routine practice. One reason for this discrepancy is most studies did not distinguish below‐the‐knee DVT from more clinically relevant above‐the‐knee DVT. Systematic reviews of medical and surgical patients have found rates of proximal propagation from 0% to 29% in untreated patients.30, 31 Though controversial, below‐the‐knee DVT is believed less morbid than proximal DVT or symptomatic DVT. We addressed this by focusing specifically on clinically relevant endpoints of proximal and symptomatic DVT. When we restricted our analysis to proximal DVT we found a 54% RR reduction in 2 pooled trials of LMWH compared to placebo. In pooled analyses symptomatic DVT was not affected by prophylaxis. When compared head‐to‐head there were no differences between LMWH and UFH for proximal DVT or symptomatic DVT.

When considering PE, the utilization of autopsy as the sole diagnostic method in 2 large trials16, 21 is particularly problematic. In the trial by Garlund,21 the mortality rate was 5.4%, with an autopsy rate of 60.1%. Similarly, in the trial by Mahe et al.,16 the mortality rate was 10%, with an autopsy rate of 49%. Given the low absolute number of deaths and substantial proportion of decedents without autopsy, the potential for chance to produce an imbalance in detection of PE is high in these studies. When we excluded these 2 trials, we found that PE was no longer reduced to a statistically significant degree by prophylaxis. Loss of significance for PE in 2 sensitivity analyses (when excluding studies of lower quality, or using autopsy as a sole diagnostic study) is problematic and calls into question the true benefit of prophylaxis for prevention of PE.

Another limitation of the current literature centers on the variability of dosing used. We pooled trials of UFH whether given BID or TID. Given the small number of trials we did not do sensitivity analyses by dosage. A recent meta‐analysis3 found both doses are efficacious, while a recent review article32 suggested superiority of TID dosing. We believe the available literature does not clearly address this issue. Regarding comparisons of LMWH to UFH, dosing variability was also noted. The trial by Bergmann and Neuhart27 used enoxaparin 20 mg per day and found similar efficacy to UFH BID, while the Samama et al.24 trial found enoxaparin 20 mg per day no more efficacious than placebo. While the literature does not clearly define a best dose, we believe enoxaparin doses lower than 40 mg daily do not reflect the standard of care.

An additional limitation of the literature is publication bias. We assessed the possibility of publication bias by a variety of means. We did find statistical evidence of publication bias for the outcome of PE when prophylaxis was compared to control. Importantly, two meta‐analyses2, 4 on thromboembolism prophylaxis for general medicine patients suggested publication bias is present and our finding supports this conclusion. While no test for publication bias is foolproof, the best protection against publication bias, which we pursued in our study, consists of a thorough search for unpublished studies, including a search of conference proceedings, contact with experts in the field, and manufacturers of LMWH.

A final limitation of the current literature centers on risk assessment. All of the trials in this meta‐analysis included patients with an elevated level of risk. Unfortunately, risk was not clearly defined in many studies, and there was no minimum level of risk between trials. While immobility, age, and length of stay were reported for most studies, other risk factors such as personal history of thromboembolism and malignancy were not uniformly reported. Based on our analysis we are not confident our results can be extrapolated to all general medicine patients.

In conclusion, we found good evidence that pharmacologic prophylaxis significantly decreases the risk of all DVT and proximal DVT in at‐risk general medical patients. However, only LMWH was shown to prevent proximal DVT. We found inconclusive evidence that prophylaxis prevents PE. When compared directly we did not find clear superiority between UFH and LMWH, though several limitations of the current literature hamper decision‐making. Given the lower cost, it may seem justified to use UFH. However, there are other practical issues, such as the fact that LMWH is given once daily, and so potentially preferred by patients and more efficient for nurses. All of these results pertain to patients with elevated risk. While we did not find significant safety concerns with prophylaxis we do not know if these results can be extrapolated to lower‐risk patients. We believe that recommending widespread prophylaxis of all general medicine patients requires additional evidence about appropriate patient selection.

Deep venous thrombosis (DVT) and pulmonary embolism (PE), collectively referred to as venous thromboembolism (VTE), are common events in hospitalized patients and result in significant morbidity and mortality. Often silent and frequently unexpected, VTE is preventable. Accordingly, the American College of Chest Physicians recommends that pharmacologic prophylaxis be given to acutely ill medical patients admitted to the hospital with congestive heart failure or severe respiratory disease, or to patients who are confined to bed who have additional risk factors, such as cancer or previous VTE.1 Three recent meta‐analyses24 demonstrated significant reductions in VTE in general medicine patients with pharmacologic prophylaxis. Recently the National Quality Forum advocated that hospitals evaluate each patient upon admission and regularly thereafter, for the risk of developing DVT/VTE and utilize clinically appropriate methods to prevent DVT/VTE.5

Despite recommendations for prophylaxis, multiple studies demonstrate utilization in <50% of at‐risk general medical patients.68 Physicians' lack of awareness may partially explain this underutilization, but other likely factors include physicians' questions about the clinical importance of the outcome (eg, some studies have shown reductions primarily in asymptomatic distal DVT), doubt regarding the best form of prophylaxis (ie, unfractionated heparin [UFH] vs. low molecular weight heparin [LMWH]), uncertainty regarding optimal dosing regimens, and comparable uncertainty regarding which patients have sufficiently high risk for VTE to outweigh the risks of anticoagulation.

We undertook the current meta‐analysis to address questions about thromboembolism prevention in general medicine patients. Does pharmacologic prophylaxis prevent clinically relevant events? Is LMWH or UFH preferable in terms of either efficacy or safety?

MATERIALS AND METHODS

RESULTS

Study Identification and Selection

The computerized literature search resulted in 5284 articles. Three additional citations were found by review of bibliographies. No additional trials were identified from reviews of abstracts from national meetings. Representatives from the 3 pharmaceutical companies reported no knowledge of additional published or unpublished data. Of the 5287 studies identified by the search, 14 studies met all eligibility criteria (Figure 1).

Figure 1

UFH or LMWH vs. Control

DVT

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of all DVT (RR = 0.55; 95% CI: 0.36‐0.83) (Figure 2A). When stratified by methodological quality, 5 trials16, 2225 with Jadad scores of 3 or higher showed an RR reduction of 0.53 (95% CI: 0.38‐0.72) in reducing all DVT. All of the higher‐quality trials compared LMWH to placebo. Across 4 trials that reported data for symptomatic DVT there was a nonsignificant reduction in RR compared with placebo (RR = 0.73; 95% CI: 0.45‐1.16) (Figure 2B). Only 2 trials24, 25 (both LMWH trials) reported results for proximal DVT and demonstrated significant benefit of prophylaxis with a pooled RR of 0.46 (95% CI: 0.31‐0.69) (Figure 2C).

Figure 2

PE

Across 7 trials comparing either UFH or LMWH to control, heparin products significantly decreased the risk of PE (RR = 0.70; 95% CI: 0.53‐0.93) (Figure 3A). The 5 trials16, 2225 with Jadad scores of 3 or greater showed a similar relative risk reduction, but the result was no longer statistically significant (RR = 0.56; 95% CI: 0.31‐1.02). Two of the trials16, 21 relied solely on the results of autopsy to diagnose PE, which may have given rise to chance differences in detection due to generally low autopsy rates. Eliminating these 2 studies from the analysis resulted in loss of statistical significance for the reduction in risk for PE (RR = 0.48; 95% CI: 0.20‐1.15).

Figure 3

Death

Seven trials16, 2025 comparing either UFH or LMWH to control examined the impact of pharmacologic prophylaxis on death and found no significant difference between treated and untreated patients across all trials (RR = 0.92; 95% CI: 0.82‐1.03) and those limited to studies with Jadad scores of 3 or higher (RR = 0.97; 95% CI: 0.80‐1.17).

Complications

We evaluated adverse events of heparin products used for prophylaxis and whether there were differences between UFH and LMWH. Reporting of complications was not uniform from study to study, making pooling more difficult. However, we were able to abstract data on any bleeding, major bleeding, and thrombocytopenia from several studies. In 5 studies15, 16, 2325 of either UFH or LMWH vs. control, a significantly increased risk of any bleeding (RR = 1.54; 95% CI: 1.15‐2.06) (Figure 4A) was found. When only major bleeding was evaluated, no statistically significant difference was noted (RR = 1.20; 95% CI: 0.55‐2.58) (Figure 4B). In 4 trials16, 22, 24, 25 the occurrence of thrombocytopenia was not significantly different when comparing UFH or LMWH to control (RR = 0.92; 95% CI: 0.46‐1.86).

Figure 4

When LMWH was compared to UFH in 4 trials,14, 17, 18, 27 a nonsignificant trend toward a decrease in any bleeding was found in the LMWH group (RR = 0.72; 95% CI: 0.44‐1.16) (Figure 5A). A similar trend was seen favoring LMWH in rates of major bleeding (RR = 0.57; 95% CI: 0.25‐1.32) (Figure 5B). Neither trend was statistically significant. Three trials comparing LMWH to UFH reported on thrombocytopenia17, 18, 27 with no significant difference noted (RR = 0.52; 95% CI: 0.06‐4.18).

Figure 5

DISCUSSION

When compared to control, LMWH or UFH decreased the risk of all DVT by 45% (RR = 0.55; 95% CI: 0.36‐0.83) and proximal DVT by 54% (RR = 0.46; 95% CI: 0.31‐0.69). PE was also decreased by 30% (RR = 0.70; 95% CI: 0.53‐0.93). Of note, when prophylaxis was compared with placebo all of the high‐quality studies showing a benefit were done using LMWH. The benefits of prophylaxis occurred at the cost of a 54% increased overall risk of bleeding (RR = 1.54; 95% CI 1.15‐2.06). However, the risk of major bleeding was not significantly increased. We did not find a mortality benefit to pharmacologic thromboembolism prophylaxis.

When comparing UFH to LMWH, we noted no difference in all DVT, symptomatic DVT, proximal DVT, PE, or death. While there was a trend toward less bleeding with LMWH, this was not statistically significant.

Taken in aggregate, our findings are in agreement with previous published meta‐analyses reporting net benefit for thromboembolism prophylaxis in medical patients.24, 22, 28, 29 Our meta‐analysis has several methodological strengths over the prior studies, including a comprehensive search of both the published and unpublished literature and assessment of the relationship between methodological quality of included trials and reported benefit. In contrast to previous reviews, our analysis highlights several limitations of the current evidence.

First, many of the studies are older, with predicted lengths of stay of greater than 1 week. The 8‐13‐day range of treatment duration we found in this study is longer than the average length of stay in today's hospitals. Second, there is variability in the diagnostic tests used to diagnose DVT, as well as variation in the definition of DVT among studies. Studies using fibrinogen uptake scanning reported rates of DVT as high as 26%15 while studies using venography reported DVT rates of almost 15% in the placebo arm.24 These rates are higher than most physicians' routine practice. One reason for this discrepancy is most studies did not distinguish below‐the‐knee DVT from more clinically relevant above‐the‐knee DVT. Systematic reviews of medical and surgical patients have found rates of proximal propagation from 0% to 29% in untreated patients.30, 31 Though controversial, below‐the‐knee DVT is believed less morbid than proximal DVT or symptomatic DVT. We addressed this by focusing specifically on clinically relevant endpoints of proximal and symptomatic DVT. When we restricted our analysis to proximal DVT we found a 54% RR reduction in 2 pooled trials of LMWH compared to placebo. In pooled analyses symptomatic DVT was not affected by prophylaxis. When compared head‐to‐head there were no differences between LMWH and UFH for proximal DVT or symptomatic DVT.

When considering PE, the utilization of autopsy as the sole diagnostic method in 2 large trials16, 21 is particularly problematic. In the trial by Garlund,21 the mortality rate was 5.4%, with an autopsy rate of 60.1%. Similarly, in the trial by Mahe et al.,16 the mortality rate was 10%, with an autopsy rate of 49%. Given the low absolute number of deaths and substantial proportion of decedents without autopsy, the potential for chance to produce an imbalance in detection of PE is high in these studies. When we excluded these 2 trials, we found that PE was no longer reduced to a statistically significant degree by prophylaxis. Loss of significance for PE in 2 sensitivity analyses (when excluding studies of lower quality, or using autopsy as a sole diagnostic study) is problematic and calls into question the true benefit of prophylaxis for prevention of PE.

Another limitation of the current literature centers on the variability of dosing used. We pooled trials of UFH whether given BID or TID. Given the small number of trials we did not do sensitivity analyses by dosage. A recent meta‐analysis3 found both doses are efficacious, while a recent review article32 suggested superiority of TID dosing. We believe the available literature does not clearly address this issue. Regarding comparisons of LMWH to UFH, dosing variability was also noted. The trial by Bergmann and Neuhart27 used enoxaparin 20 mg per day and found similar efficacy to UFH BID, while the Samama et al.24 trial found enoxaparin 20 mg per day no more efficacious than placebo. While the literature does not clearly define a best dose, we believe enoxaparin doses lower than 40 mg daily do not reflect the standard of care.

An additional limitation of the literature is publication bias. We assessed the possibility of publication bias by a variety of means. We did find statistical evidence of publication bias for the outcome of PE when prophylaxis was compared to control. Importantly, two meta‐analyses2, 4 on thromboembolism prophylaxis for general medicine patients suggested publication bias is present and our finding supports this conclusion. While no test for publication bias is foolproof, the best protection against publication bias, which we pursued in our study, consists of a thorough search for unpublished studies, including a search of conference proceedings, contact with experts in the field, and manufacturers of LMWH.

A final limitation of the current literature centers on risk assessment. All of the trials in this meta‐analysis included patients with an elevated level of risk. Unfortunately, risk was not clearly defined in many studies, and there was no minimum level of risk between trials. While immobility, age, and length of stay were reported for most studies, other risk factors such as personal history of thromboembolism and malignancy were not uniformly reported. Based on our analysis we are not confident our results can be extrapolated to all general medicine patients.

In conclusion, we found good evidence that pharmacologic prophylaxis significantly decreases the risk of all DVT and proximal DVT in at‐risk general medical patients. However, only LMWH was shown to prevent proximal DVT. We found inconclusive evidence that prophylaxis prevents PE. When compared directly we did not find clear superiority between UFH and LMWH, though several limitations of the current literature hamper decision‐making. Given the lower cost, it may seem justified to use UFH. However, there are other practical issues, such as the fact that LMWH is given once daily, and so potentially preferred by patients and more efficient for nurses. All of these results pertain to patients with elevated risk. While we did not find significant safety concerns with prophylaxis we do not know if these results can be extrapolated to lower‐risk patients. We believe that recommending widespread prophylaxis of all general medicine patients requires additional evidence about appropriate patient selection.

A 78‐year‐old man was transferred from an outside hospital where he presented with declining mental status and a history of falls. A computed tomography (CT) scan of the brain revealed a chronic subdural hematoma with superimposed acute hemorrhage. The subdural hematoma was attributed to a fall at home approximately 5 weeks prior to admission. He was taken to the operating room for urgent craniotomy and hemorrhage evacuation and postoperatively comanaged by neurosurgery, hospitalists, and medicine residents. He tolerated the procedure and was noted to have marked improvement in mental status after the procedure. He was monitored overnight in our intensive care unit without intracranial pressure monitoring.

Early on postoperative day 1, he was awake, alert, following commands, and felt to be stable enough to be transferred to our transitional intensive care unit. However, later in the day he became progressively more confused. A follow‐up CT scan of the brain was ordered (Fig. 1) by the medicine team which revealed a large collection of air (wide arrow) and marked midline shift (thin arrow) consistent with tension pneumocephaly and subfalcine herniation (Fig. 2; arrow). Examination revealed that he was grossly obtunded with marked anisocoria, decerebrate posturing, and rigid tone. Neurosurgery was immediately contacted and recommended accessing 2 indwelling catheters left in the cerebrum as part of the normal postoperative course. Approximately 100 mL of serosanguinous fluid and air was aspirated with immediate improvement in his mental status and exam findings. Over the next few days, he remained clinically stable, and repeat CT scan showed slow resolution of the pneumocephalus and a decrease of his mass effect and midline shift. He was ultimately transferred to our skilled nursing facility for physical therapy and has done relatively well.

Figure 1

Figure 2

Pneumocephalus is a relatively common finding in many neurosurgical, intracranial procedures. However, tension pneumocephalus is a rare, life‐threatening form of pneumocephalus in which intracranial air causes mass effect and midline shift. In a review of 295 cases of pneumocephalus, 75% were caused by surgery, mostly intracranial and transsphenoidal, and head trauma. About 9% of cases resulted from infection with gas‐forming bacteria and rare causes include invasion of a nasopharyngeal carcinoma, frequent Valsalva maneuver, and air travel.1 Tension pneumocephalus occurs most commonly after the neurosurgical evacuation of a subdural hematoma. The prevalence of tension pneumocephalus following the evacuation of chronic subdural hematomas has been reported from 2.5% to 16%.2

There are 2 proposed mechanisms for the development of pneumocephalus; 1 proposes that air passes through a dural tear by a ball valve effect in which air can be forced into the intracranial cavity by a rapid increase in intrasinus pressure that occurs during sneezing, coughing, or straining. The air is then trapped intracranially. The second theory proposes that cerebrospinal fluid leakage permits air to enter the intracranial cavity because negative pressure is created as cerebrospinal fluid leaves the space.3 The conversion to tension physiology in either of these theories is less well understood.

A 78‐year‐old man was transferred from an outside hospital where he presented with declining mental status and a history of falls. A computed tomography (CT) scan of the brain revealed a chronic subdural hematoma with superimposed acute hemorrhage. The subdural hematoma was attributed to a fall at home approximately 5 weeks prior to admission. He was taken to the operating room for urgent craniotomy and hemorrhage evacuation and postoperatively comanaged by neurosurgery, hospitalists, and medicine residents. He tolerated the procedure and was noted to have marked improvement in mental status after the procedure. He was monitored overnight in our intensive care unit without intracranial pressure monitoring.

Early on postoperative day 1, he was awake, alert, following commands, and felt to be stable enough to be transferred to our transitional intensive care unit. However, later in the day he became progressively more confused. A follow‐up CT scan of the brain was ordered (Fig. 1) by the medicine team which revealed a large collection of air (wide arrow) and marked midline shift (thin arrow) consistent with tension pneumocephaly and subfalcine herniation (Fig. 2; arrow). Examination revealed that he was grossly obtunded with marked anisocoria, decerebrate posturing, and rigid tone. Neurosurgery was immediately contacted and recommended accessing 2 indwelling catheters left in the cerebrum as part of the normal postoperative course. Approximately 100 mL of serosanguinous fluid and air was aspirated with immediate improvement in his mental status and exam findings. Over the next few days, he remained clinically stable, and repeat CT scan showed slow resolution of the pneumocephalus and a decrease of his mass effect and midline shift. He was ultimately transferred to our skilled nursing facility for physical therapy and has done relatively well.

Figure 1

Figure 2

Pneumocephalus is a relatively common finding in many neurosurgical, intracranial procedures. However, tension pneumocephalus is a rare, life‐threatening form of pneumocephalus in which intracranial air causes mass effect and midline shift. In a review of 295 cases of pneumocephalus, 75% were caused by surgery, mostly intracranial and transsphenoidal, and head trauma. About 9% of cases resulted from infection with gas‐forming bacteria and rare causes include invasion of a nasopharyngeal carcinoma, frequent Valsalva maneuver, and air travel.1 Tension pneumocephalus occurs most commonly after the neurosurgical evacuation of a subdural hematoma. The prevalence of tension pneumocephalus following the evacuation of chronic subdural hematomas has been reported from 2.5% to 16%.2

There are 2 proposed mechanisms for the development of pneumocephalus; 1 proposes that air passes through a dural tear by a ball valve effect in which air can be forced into the intracranial cavity by a rapid increase in intrasinus pressure that occurs during sneezing, coughing, or straining. The air is then trapped intracranially. The second theory proposes that cerebrospinal fluid leakage permits air to enter the intracranial cavity because negative pressure is created as cerebrospinal fluid leaves the space.3 The conversion to tension physiology in either of these theories is less well understood.

A 78‐year‐old man was transferred from an outside hospital where he presented with declining mental status and a history of falls. A computed tomography (CT) scan of the brain revealed a chronic subdural hematoma with superimposed acute hemorrhage. The subdural hematoma was attributed to a fall at home approximately 5 weeks prior to admission. He was taken to the operating room for urgent craniotomy and hemorrhage evacuation and postoperatively comanaged by neurosurgery, hospitalists, and medicine residents. He tolerated the procedure and was noted to have marked improvement in mental status after the procedure. He was monitored overnight in our intensive care unit without intracranial pressure monitoring.

Early on postoperative day 1, he was awake, alert, following commands, and felt to be stable enough to be transferred to our transitional intensive care unit. However, later in the day he became progressively more confused. A follow‐up CT scan of the brain was ordered (Fig. 1) by the medicine team which revealed a large collection of air (wide arrow) and marked midline shift (thin arrow) consistent with tension pneumocephaly and subfalcine herniation (Fig. 2; arrow). Examination revealed that he was grossly obtunded with marked anisocoria, decerebrate posturing, and rigid tone. Neurosurgery was immediately contacted and recommended accessing 2 indwelling catheters left in the cerebrum as part of the normal postoperative course. Approximately 100 mL of serosanguinous fluid and air was aspirated with immediate improvement in his mental status and exam findings. Over the next few days, he remained clinically stable, and repeat CT scan showed slow resolution of the pneumocephalus and a decrease of his mass effect and midline shift. He was ultimately transferred to our skilled nursing facility for physical therapy and has done relatively well.

Figure 1

Figure 2

Pneumocephalus is a relatively common finding in many neurosurgical, intracranial procedures. However, tension pneumocephalus is a rare, life‐threatening form of pneumocephalus in which intracranial air causes mass effect and midline shift. In a review of 295 cases of pneumocephalus, 75% were caused by surgery, mostly intracranial and transsphenoidal, and head trauma. About 9% of cases resulted from infection with gas‐forming bacteria and rare causes include invasion of a nasopharyngeal carcinoma, frequent Valsalva maneuver, and air travel.1 Tension pneumocephalus occurs most commonly after the neurosurgical evacuation of a subdural hematoma. The prevalence of tension pneumocephalus following the evacuation of chronic subdural hematomas has been reported from 2.5% to 16%.2

There are 2 proposed mechanisms for the development of pneumocephalus; 1 proposes that air passes through a dural tear by a ball valve effect in which air can be forced into the intracranial cavity by a rapid increase in intrasinus pressure that occurs during sneezing, coughing, or straining. The air is then trapped intracranially. The second theory proposes that cerebrospinal fluid leakage permits air to enter the intracranial cavity because negative pressure is created as cerebrospinal fluid leaves the space.3 The conversion to tension physiology in either of these theories is less well understood.