Associate Editor In Chief

Missed Cervical Spine Injury

An 83-year-old man presented to the ED via emergency medical services (EMS) with a chief complaint of neck pain. He was the restrained driver of a car that was struck from behind by another vehicle. The patient denied any head injury, loss of consciousness, chest pain, shortness of breath, or abdominal pain. His medical history was significant for hypertension and coronary artery disease, for which he was taking several medications. Regarding his social history, the patient denied alcohol consumption or cigarette smoking.

The patient’s physical examination was unremarkable. His vital signs were normal, and there was no obvious external evidence of trauma. The posterior cervical spine was tender to palpation in the midline, but no step-off signs were appreciated. The neurological examination, including strength and sensation in all four extremities, was normal.

Since the patient’s only complaint was neck pain and his physical examination and history were otherwise normal, the emergency physician (EP) ordered radiographs of the cervical spine. The imaging studies were interpreted as showing advanced degenerative changes but no fractures, and the patient was prescribed an analgesic and discharged home.

When the patient woke up the next morning, he was unable to move his extremities, and returned to the same ED via EMS. He was placed in a cervical collar and found to have flaccid extremities on examination. A computed tomography (CT) scan of the cervical spine revealed a transverse fracture through the C6 vertebra. Radiology services also reviewed the cervical spine X-rays from the previous day, noting the presence of fracture.

The patient was taken to the operating room by neurosurgery services but remained paralyzed postoperatively. He never recovered from his injury and died 6 months later. His family sued the EP and the hospital for missed diagnosis of cervical spine fracture at the first ED presentation and the resulting paralysis. The case was settled for $1.3 million prior to trial.

Discussion

The evaluation of suspected cervical spine injury secondary to blunt trauma is a frequent and important skill practiced by EPs. Motor vehicle accidents are the most common cause of spinal cord injury in the United States (42%), followed by falls (27%), acts of violence (15%), and sports-related injuries (8%).¹ A review by Sekon and Fehlings² showed that 55% of all spinal injuries involve the cervical spine. Interestingly, the majority of cervical spine injuries occur at the upper or lower ends of the cervical spine; C2 vertebral fractures account for 33%, while C6 and C7 vertebral fractures account for approximately 50%.¹

There are two commonly used criteria to clinically clear the cervical spine (ie, no imaging studies necessary) in blunt-trauma patients. The first is the National Emergency X-Radiography Use Study (NEXUS), which has a sensitivity of 99.6% of identifying cervical spine fractures.¹ According to the NEXUS criteria, no imaging studies are required if: (1) there is no midline cervical spine tenderness; (2) there are no focal neurological deficits; (3) the patient exhibits a normal level of alertness; (4) the patient is not intoxicated; and (5) there is no distracting injury.¹

The other set of criteria used to clear the cervical spine is the Canadian Cervical Spine Rule. In these criteria, a patient is considered at very low risk for cervical spine fracture in the following cases: (1) the patient is fully alert with a Glasgow Coma scale of 15; (2) the patient has no high-risk factors (ie, age >65 years, dangerous mechanism of injury, fall greater than five stairs, axial load to the head, high-speed vehicular crash, bicycle or motorcycle crash, or the presence of paresthesias in the extremities); (3) the patient has low-risk factors (eg, simple vehicle crash, sitting position in the ED, ambulatory at any time, delayed onset of neck pain, and the absence of midline cervical tenderness); and (4) the patient can actively rotate his or her neck 45 degrees to the left and to the right. The Canadian group found the above criteria to have 100% sensitivity for predicting the absence of cervical spine injury.¹

The patient in this case failed both sets of criteria (ie, presence of cervical spine tenderness and age >65 years) and therefore required imaging. Historically, cervical spine X-ray (three views, anteroposterior, lateral, and odontoid; or five views, three views plus obliques) has been the imaging study of choice for such patients. Unfortunately, however, cervical spine radiographs have severe limitations in identifying spinal injury. In a large retrospective review, Woodring and Lee,³ found that the standard three-view cervical spine series failed to demonstrate 61% of all fractures and 36% of all subluxation and dislocations. Similarly, in a prospective study of 1,006 patients with 72 injuries, Diaz et al,⁴ found a 52.3% missed fracture rate when five-view radiographs were used to identify cervical spine injury. In addition, radiographic evaluation of elderly patients was found to be even more challenging in identifying cervical spine injury due to age-related degenerative changes.

 

Given the abovementioned limitations associated with radiographic imaging, CT scan of the cervical spine has become the imaging study of choice in moderate-to-severe risk patients with blunt cervical spine trauma. This modality has been shown to have a higher sensitivity and specificity for evaluating cervical spine injury compared to plain X-ray films, with CT detecting 97% to 100% of cervical spine fractures.⁵

In addition to demonstrating a higher sensitivity, CT also has the advantage of speed—especially when the patient is undergoing other CT studies (eg, head, abdomen, pelvis). While some clinicians criticize the higher cost of CT versus plain films, CT has been shown to decrease institutional costs (when settlement costs are taken into account) due to the reduction of the incidence of paralysis resulting from false-negative imaging studies.⁶

Forgotten Tourniquet

A 33-year-old woman presented to the ED with a chief complaint of left-sided abdominal and flank pain. She described the onset of pain as abrupt, severe, and lasting approximately 3 hours in duration. She admitted to nausea, but no vomiting. She also denied a history of any previous similar symptoms or recent trauma. The patient’s medical history was unremarkable. Her last menstrual period began 3 days prior to presentation. Regarding social history, she denied any tobacco or alcohol use.

The patient’s vital signs were: blood pressure, 138/82 mm Hg; heart rate, 102 beats/minute; respiratory rate, 18 breaths/minute; temperature 98.6˚F. Oxygen saturation was 99% on room air.

The patient appeared uncomfortable overall. The physical examination was remarkable only for mild left-sided costovertebral angle tenderness. Her abdomen was soft, nontender, and without guarding or rebound.

The EP ordered the placement of an intravenous (IV) line, through which the patient was administered normal saline and morphine and promethazine, respectively, for pain and nausea. A complete blood count, basic metabolic panel, urinalysis, and urine pregnancy test were ordered. All of the laboratory bloodwork results were normal, and the urine pregnancy test was negative. The urinalysis was remarkable for 50 to 100 red blood cells.

A noncontrast CT scan of the abdomen and pelvis revealed a 3-mm ureteral stone on the left side. When the patient returned from radiology services, her pain was significantly decreased and she felt much improved. She was diagnosed with a kidney stone and discharged home with an analgesic and a strainer, along with instructions to follow-up with urology services. The patient was in the ED for a total of 5 hours.

The plaintiff sued the EP and hospital, claiming that the tourniquet used to start the IV line and draw blood was never removed, which in turn caused nerve damage resulting in reflex sympathetic dystrophy and complex regional pain syndrome. The defense denied all of these allegations, and the ED personnel testified that the tourniquet was removed as soon as the IV was established. The defense cited the plaintiff’s medical records, which contained documentation that the tourniquet had been removed. The defense further argued that if the tourniquet had been left on as the patient alleged, she would have experienced obvious physical signs, such as swelling, redness, infiltration of fluids, pain, and numbness. A defense verdict was returned.

Discussion

It is very tempting to simply dismiss this case as absurd, with nothing to be learned from it. It does defy common sense that no one would have noticed the tourniquet or, at the very least, that the patient would not have spoken up about it during her stay in the ED. While the jury clearly came to the correct conclusion,  it does highlight a real problem: forgotten tourniquets.

According to the Pennsylvania Patient Safety Advisory (PPSA), there were 125 reports of tourniquets being left on patients in Pennsylvania healthcare facilities in 1 year alone.¹ In 5% of these cases, the tourniquet was discovered within a half hour of application. In approximately 66% of cases, the tourniquet was left on for up to 2 hours, and the remaining were left in place for 2 to 18 hours.

Few locations within the hospital are without risk for this type of accident. The PPSA further noted that approximately 30% of retained tourniquets occurred on medical/surgical units, 14% in the ED, and 14% on inpatient and ambulatory surgical services departments. Approximately 19% were discovered when patients were transferred from one department to another.¹

In the analysis of these incidents, contributing factors to forgotten tourniquets included staff failing to follow proper procedures, inadequate staff proficiency, and staff distractions and/or interruptions.¹ In addition, some patients appeared to be at increased risk of having a retained tourniquet than others. Sixty percent of 125 patients with a forgotten tourniquet were aged 70 years or older, whereas some patients were younger than age 2 years.¹ Not surprisingly, patients who were unable to verbally communicate (eg, patients who were intubated, under anesthesia, had expressive aphasia, severe dementia), were at the highest risk.

 

In a review of recovery room incidents, Salman and Asfar² identified two cases of forgotten tourniquets out of approximately 7,000 patients. Potential strategies to avoid this mistake include: (1) only documenting procedures after they have been completed (eg, tourniquet removal); (2) double-checking that the tourniquet has been removed prior to leaving patient bedside; and (3) the use of extra-long tourniquets so the ends are more clearly visible.

Missed Cervical Spine Injury

An 83-year-old man presented to the ED via emergency medical services (EMS) with a chief complaint of neck pain. He was the restrained driver of a car that was struck from behind by another vehicle. The patient denied any head injury, loss of consciousness, chest pain, shortness of breath, or abdominal pain. His medical history was significant for hypertension and coronary artery disease, for which he was taking several medications. Regarding his social history, the patient denied alcohol consumption or cigarette smoking.

The patient’s physical examination was unremarkable. His vital signs were normal, and there was no obvious external evidence of trauma. The posterior cervical spine was tender to palpation in the midline, but no step-off signs were appreciated. The neurological examination, including strength and sensation in all four extremities, was normal.

Since the patient’s only complaint was neck pain and his physical examination and history were otherwise normal, the emergency physician (EP) ordered radiographs of the cervical spine. The imaging studies were interpreted as showing advanced degenerative changes but no fractures, and the patient was prescribed an analgesic and discharged home.

When the patient woke up the next morning, he was unable to move his extremities, and returned to the same ED via EMS. He was placed in a cervical collar and found to have flaccid extremities on examination. A computed tomography (CT) scan of the cervical spine revealed a transverse fracture through the C6 vertebra. Radiology services also reviewed the cervical spine X-rays from the previous day, noting the presence of fracture.

The patient was taken to the operating room by neurosurgery services but remained paralyzed postoperatively. He never recovered from his injury and died 6 months later. His family sued the EP and the hospital for missed diagnosis of cervical spine fracture at the first ED presentation and the resulting paralysis. The case was settled for $1.3 million prior to trial.

Discussion

The evaluation of suspected cervical spine injury secondary to blunt trauma is a frequent and important skill practiced by EPs. Motor vehicle accidents are the most common cause of spinal cord injury in the United States (42%), followed by falls (27%), acts of violence (15%), and sports-related injuries (8%).¹ A review by Sekon and Fehlings² showed that 55% of all spinal injuries involve the cervical spine. Interestingly, the majority of cervical spine injuries occur at the upper or lower ends of the cervical spine; C2 vertebral fractures account for 33%, while C6 and C7 vertebral fractures account for approximately 50%.¹

There are two commonly used criteria to clinically clear the cervical spine (ie, no imaging studies necessary) in blunt-trauma patients. The first is the National Emergency X-Radiography Use Study (NEXUS), which has a sensitivity of 99.6% of identifying cervical spine fractures.¹ According to the NEXUS criteria, no imaging studies are required if: (1) there is no midline cervical spine tenderness; (2) there are no focal neurological deficits; (3) the patient exhibits a normal level of alertness; (4) the patient is not intoxicated; and (5) there is no distracting injury.¹

The other set of criteria used to clear the cervical spine is the Canadian Cervical Spine Rule. In these criteria, a patient is considered at very low risk for cervical spine fracture in the following cases: (1) the patient is fully alert with a Glasgow Coma scale of 15; (2) the patient has no high-risk factors (ie, age >65 years, dangerous mechanism of injury, fall greater than five stairs, axial load to the head, high-speed vehicular crash, bicycle or motorcycle crash, or the presence of paresthesias in the extremities); (3) the patient has low-risk factors (eg, simple vehicle crash, sitting position in the ED, ambulatory at any time, delayed onset of neck pain, and the absence of midline cervical tenderness); and (4) the patient can actively rotate his or her neck 45 degrees to the left and to the right. The Canadian group found the above criteria to have 100% sensitivity for predicting the absence of cervical spine injury.¹

The patient in this case failed both sets of criteria (ie, presence of cervical spine tenderness and age >65 years) and therefore required imaging. Historically, cervical spine X-ray (three views, anteroposterior, lateral, and odontoid; or five views, three views plus obliques) has been the imaging study of choice for such patients. Unfortunately, however, cervical spine radiographs have severe limitations in identifying spinal injury. In a large retrospective review, Woodring and Lee,³ found that the standard three-view cervical spine series failed to demonstrate 61% of all fractures and 36% of all subluxation and dislocations. Similarly, in a prospective study of 1,006 patients with 72 injuries, Diaz et al,⁴ found a 52.3% missed fracture rate when five-view radiographs were used to identify cervical spine injury. In addition, radiographic evaluation of elderly patients was found to be even more challenging in identifying cervical spine injury due to age-related degenerative changes.

 

Given the abovementioned limitations associated with radiographic imaging, CT scan of the cervical spine has become the imaging study of choice in moderate-to-severe risk patients with blunt cervical spine trauma. This modality has been shown to have a higher sensitivity and specificity for evaluating cervical spine injury compared to plain X-ray films, with CT detecting 97% to 100% of cervical spine fractures.⁵

In addition to demonstrating a higher sensitivity, CT also has the advantage of speed—especially when the patient is undergoing other CT studies (eg, head, abdomen, pelvis). While some clinicians criticize the higher cost of CT versus plain films, CT has been shown to decrease institutional costs (when settlement costs are taken into account) due to the reduction of the incidence of paralysis resulting from false-negative imaging studies.⁶

Forgotten Tourniquet

A 33-year-old woman presented to the ED with a chief complaint of left-sided abdominal and flank pain. She described the onset of pain as abrupt, severe, and lasting approximately 3 hours in duration. She admitted to nausea, but no vomiting. She also denied a history of any previous similar symptoms or recent trauma. The patient’s medical history was unremarkable. Her last menstrual period began 3 days prior to presentation. Regarding social history, she denied any tobacco or alcohol use.

The patient’s vital signs were: blood pressure, 138/82 mm Hg; heart rate, 102 beats/minute; respiratory rate, 18 breaths/minute; temperature 98.6˚F. Oxygen saturation was 99% on room air.

The patient appeared uncomfortable overall. The physical examination was remarkable only for mild left-sided costovertebral angle tenderness. Her abdomen was soft, nontender, and without guarding or rebound.

The EP ordered the placement of an intravenous (IV) line, through which the patient was administered normal saline and morphine and promethazine, respectively, for pain and nausea. A complete blood count, basic metabolic panel, urinalysis, and urine pregnancy test were ordered. All of the laboratory bloodwork results were normal, and the urine pregnancy test was negative. The urinalysis was remarkable for 50 to 100 red blood cells.

A noncontrast CT scan of the abdomen and pelvis revealed a 3-mm ureteral stone on the left side. When the patient returned from radiology services, her pain was significantly decreased and she felt much improved. She was diagnosed with a kidney stone and discharged home with an analgesic and a strainer, along with instructions to follow-up with urology services. The patient was in the ED for a total of 5 hours.

The plaintiff sued the EP and hospital, claiming that the tourniquet used to start the IV line and draw blood was never removed, which in turn caused nerve damage resulting in reflex sympathetic dystrophy and complex regional pain syndrome. The defense denied all of these allegations, and the ED personnel testified that the tourniquet was removed as soon as the IV was established. The defense cited the plaintiff’s medical records, which contained documentation that the tourniquet had been removed. The defense further argued that if the tourniquet had been left on as the patient alleged, she would have experienced obvious physical signs, such as swelling, redness, infiltration of fluids, pain, and numbness. A defense verdict was returned.

Discussion

It is very tempting to simply dismiss this case as absurd, with nothing to be learned from it. It does defy common sense that no one would have noticed the tourniquet or, at the very least, that the patient would not have spoken up about it during her stay in the ED. While the jury clearly came to the correct conclusion,  it does highlight a real problem: forgotten tourniquets.

According to the Pennsylvania Patient Safety Advisory (PPSA), there were 125 reports of tourniquets being left on patients in Pennsylvania healthcare facilities in 1 year alone.¹ In 5% of these cases, the tourniquet was discovered within a half hour of application. In approximately 66% of cases, the tourniquet was left on for up to 2 hours, and the remaining were left in place for 2 to 18 hours.

Few locations within the hospital are without risk for this type of accident. The PPSA further noted that approximately 30% of retained tourniquets occurred on medical/surgical units, 14% in the ED, and 14% on inpatient and ambulatory surgical services departments. Approximately 19% were discovered when patients were transferred from one department to another.¹

In the analysis of these incidents, contributing factors to forgotten tourniquets included staff failing to follow proper procedures, inadequate staff proficiency, and staff distractions and/or interruptions.¹ In addition, some patients appeared to be at increased risk of having a retained tourniquet than others. Sixty percent of 125 patients with a forgotten tourniquet were aged 70 years or older, whereas some patients were younger than age 2 years.¹ Not surprisingly, patients who were unable to verbally communicate (eg, patients who were intubated, under anesthesia, had expressive aphasia, severe dementia), were at the highest risk.

 

In a review of recovery room incidents, Salman and Asfar² identified two cases of forgotten tourniquets out of approximately 7,000 patients. Potential strategies to avoid this mistake include: (1) only documenting procedures after they have been completed (eg, tourniquet removal); (2) double-checking that the tourniquet has been removed prior to leaving patient bedside; and (3) the use of extra-long tourniquets so the ends are more clearly visible.

Missed Cervical Spine Injury

An 83-year-old man presented to the ED via emergency medical services (EMS) with a chief complaint of neck pain. He was the restrained driver of a car that was struck from behind by another vehicle. The patient denied any head injury, loss of consciousness, chest pain, shortness of breath, or abdominal pain. His medical history was significant for hypertension and coronary artery disease, for which he was taking several medications. Regarding his social history, the patient denied alcohol consumption or cigarette smoking.

The patient’s physical examination was unremarkable. His vital signs were normal, and there was no obvious external evidence of trauma. The posterior cervical spine was tender to palpation in the midline, but no step-off signs were appreciated. The neurological examination, including strength and sensation in all four extremities, was normal.

Since the patient’s only complaint was neck pain and his physical examination and history were otherwise normal, the emergency physician (EP) ordered radiographs of the cervical spine. The imaging studies were interpreted as showing advanced degenerative changes but no fractures, and the patient was prescribed an analgesic and discharged home.

When the patient woke up the next morning, he was unable to move his extremities, and returned to the same ED via EMS. He was placed in a cervical collar and found to have flaccid extremities on examination. A computed tomography (CT) scan of the cervical spine revealed a transverse fracture through the C6 vertebra. Radiology services also reviewed the cervical spine X-rays from the previous day, noting the presence of fracture.

The patient was taken to the operating room by neurosurgery services but remained paralyzed postoperatively. He never recovered from his injury and died 6 months later. His family sued the EP and the hospital for missed diagnosis of cervical spine fracture at the first ED presentation and the resulting paralysis. The case was settled for $1.3 million prior to trial.

Discussion

The evaluation of suspected cervical spine injury secondary to blunt trauma is a frequent and important skill practiced by EPs. Motor vehicle accidents are the most common cause of spinal cord injury in the United States (42%), followed by falls (27%), acts of violence (15%), and sports-related injuries (8%).¹ A review by Sekon and Fehlings² showed that 55% of all spinal injuries involve the cervical spine. Interestingly, the majority of cervical spine injuries occur at the upper or lower ends of the cervical spine; C2 vertebral fractures account for 33%, while C6 and C7 vertebral fractures account for approximately 50%.¹

There are two commonly used criteria to clinically clear the cervical spine (ie, no imaging studies necessary) in blunt-trauma patients. The first is the National Emergency X-Radiography Use Study (NEXUS), which has a sensitivity of 99.6% of identifying cervical spine fractures.¹ According to the NEXUS criteria, no imaging studies are required if: (1) there is no midline cervical spine tenderness; (2) there are no focal neurological deficits; (3) the patient exhibits a normal level of alertness; (4) the patient is not intoxicated; and (5) there is no distracting injury.¹

The other set of criteria used to clear the cervical spine is the Canadian Cervical Spine Rule. In these criteria, a patient is considered at very low risk for cervical spine fracture in the following cases: (1) the patient is fully alert with a Glasgow Coma scale of 15; (2) the patient has no high-risk factors (ie, age >65 years, dangerous mechanism of injury, fall greater than five stairs, axial load to the head, high-speed vehicular crash, bicycle or motorcycle crash, or the presence of paresthesias in the extremities); (3) the patient has low-risk factors (eg, simple vehicle crash, sitting position in the ED, ambulatory at any time, delayed onset of neck pain, and the absence of midline cervical tenderness); and (4) the patient can actively rotate his or her neck 45 degrees to the left and to the right. The Canadian group found the above criteria to have 100% sensitivity for predicting the absence of cervical spine injury.¹

The patient in this case failed both sets of criteria (ie, presence of cervical spine tenderness and age >65 years) and therefore required imaging. Historically, cervical spine X-ray (three views, anteroposterior, lateral, and odontoid; or five views, three views plus obliques) has been the imaging study of choice for such patients. Unfortunately, however, cervical spine radiographs have severe limitations in identifying spinal injury. In a large retrospective review, Woodring and Lee,³ found that the standard three-view cervical spine series failed to demonstrate 61% of all fractures and 36% of all subluxation and dislocations. Similarly, in a prospective study of 1,006 patients with 72 injuries, Diaz et al,⁴ found a 52.3% missed fracture rate when five-view radiographs were used to identify cervical spine injury. In addition, radiographic evaluation of elderly patients was found to be even more challenging in identifying cervical spine injury due to age-related degenerative changes.

 

Given the abovementioned limitations associated with radiographic imaging, CT scan of the cervical spine has become the imaging study of choice in moderate-to-severe risk patients with blunt cervical spine trauma. This modality has been shown to have a higher sensitivity and specificity for evaluating cervical spine injury compared to plain X-ray films, with CT detecting 97% to 100% of cervical spine fractures.⁵

In addition to demonstrating a higher sensitivity, CT also has the advantage of speed—especially when the patient is undergoing other CT studies (eg, head, abdomen, pelvis). While some clinicians criticize the higher cost of CT versus plain films, CT has been shown to decrease institutional costs (when settlement costs are taken into account) due to the reduction of the incidence of paralysis resulting from false-negative imaging studies.⁶

Forgotten Tourniquet

A 33-year-old woman presented to the ED with a chief complaint of left-sided abdominal and flank pain. She described the onset of pain as abrupt, severe, and lasting approximately 3 hours in duration. She admitted to nausea, but no vomiting. She also denied a history of any previous similar symptoms or recent trauma. The patient’s medical history was unremarkable. Her last menstrual period began 3 days prior to presentation. Regarding social history, she denied any tobacco or alcohol use.

The patient’s vital signs were: blood pressure, 138/82 mm Hg; heart rate, 102 beats/minute; respiratory rate, 18 breaths/minute; temperature 98.6˚F. Oxygen saturation was 99% on room air.

The patient appeared uncomfortable overall. The physical examination was remarkable only for mild left-sided costovertebral angle tenderness. Her abdomen was soft, nontender, and without guarding or rebound.

The EP ordered the placement of an intravenous (IV) line, through which the patient was administered normal saline and morphine and promethazine, respectively, for pain and nausea. A complete blood count, basic metabolic panel, urinalysis, and urine pregnancy test were ordered. All of the laboratory bloodwork results were normal, and the urine pregnancy test was negative. The urinalysis was remarkable for 50 to 100 red blood cells.

A noncontrast CT scan of the abdomen and pelvis revealed a 3-mm ureteral stone on the left side. When the patient returned from radiology services, her pain was significantly decreased and she felt much improved. She was diagnosed with a kidney stone and discharged home with an analgesic and a strainer, along with instructions to follow-up with urology services. The patient was in the ED for a total of 5 hours.

The plaintiff sued the EP and hospital, claiming that the tourniquet used to start the IV line and draw blood was never removed, which in turn caused nerve damage resulting in reflex sympathetic dystrophy and complex regional pain syndrome. The defense denied all of these allegations, and the ED personnel testified that the tourniquet was removed as soon as the IV was established. The defense cited the plaintiff’s medical records, which contained documentation that the tourniquet had been removed. The defense further argued that if the tourniquet had been left on as the patient alleged, she would have experienced obvious physical signs, such as swelling, redness, infiltration of fluids, pain, and numbness. A defense verdict was returned.

Discussion

It is very tempting to simply dismiss this case as absurd, with nothing to be learned from it. It does defy common sense that no one would have noticed the tourniquet or, at the very least, that the patient would not have spoken up about it during her stay in the ED. While the jury clearly came to the correct conclusion,  it does highlight a real problem: forgotten tourniquets.

According to the Pennsylvania Patient Safety Advisory (PPSA), there were 125 reports of tourniquets being left on patients in Pennsylvania healthcare facilities in 1 year alone.¹ In 5% of these cases, the tourniquet was discovered within a half hour of application. In approximately 66% of cases, the tourniquet was left on for up to 2 hours, and the remaining were left in place for 2 to 18 hours.

Few locations within the hospital are without risk for this type of accident. The PPSA further noted that approximately 30% of retained tourniquets occurred on medical/surgical units, 14% in the ED, and 14% on inpatient and ambulatory surgical services departments. Approximately 19% were discovered when patients were transferred from one department to another.¹

In the analysis of these incidents, contributing factors to forgotten tourniquets included staff failing to follow proper procedures, inadequate staff proficiency, and staff distractions and/or interruptions.¹ In addition, some patients appeared to be at increased risk of having a retained tourniquet than others. Sixty percent of 125 patients with a forgotten tourniquet were aged 70 years or older, whereas some patients were younger than age 2 years.¹ Not surprisingly, patients who were unable to verbally communicate (eg, patients who were intubated, under anesthesia, had expressive aphasia, severe dementia), were at the highest risk.

 

In a review of recovery room incidents, Salman and Asfar² identified two cases of forgotten tourniquets out of approximately 7,000 patients. Potential strategies to avoid this mistake include: (1) only documenting procedures after they have been completed (eg, tourniquet removal); (2) double-checking that the tourniquet has been removed prior to leaving patient bedside; and (3) the use of extra-long tourniquets so the ends are more clearly visible.

Missed Preeclampsia

A 24-year-old woman, gravida 1, para 1, aborta 0, presented to the ED complaining of a 1-day history of shortness of breath. Four days earlier, she had delivered a healthy baby boy via normal vaginal delivery and without complication. She denied chest pain, fever, or abdominal pain. She was otherwise in good health, stating that she was not taking any medications. She also denied smoking cigarettes.

On physical examination, the patient’s vital signs were remarkable for the following: heart rate (HR), 86 beats/minute; blood pressure (BP), 164/94 mm Hg; respiratory rate, 18 breaths/minute; temperature, 98.6^oF. Oxygen saturation was 96% on room air. The head, eye, ear, nose and throat examination was unremarkable. The lungs were clear to auscultation bilaterally, and HR and heart rhythm were normal. The abdomen was soft and nontender without guarding or rebound. The lower extremities were remarkable for 1+ pedal and pretibial edema bilaterally.

Since this patient was 4 days postpartum, the emergency physician (EP) was concerned for pulmonary embolism (PE). A complete blood count, basic metabolic profile, and a serum troponin T level were ordered. The electrocardiogram revealed normal sinus rhythm without evidence of strain or injury. The chest X-ray was interpreted by radiology services as normal. Given the concern for PE, computed tomography angiography (CTA) of the chest was ordered. All laboratory studies, including the troponin T level, were reported as normal. The CTA scan of the chest was interpreted by radiology services as normal and without evidence of PE. The patient was discharged home with a diagnosis of “shortness of breath of unknown etiology.”

The patient presented to the same ED 2 days later, again with the chief complaint of shortness of breath. On examination, her BP was noted to be elevated and she had 1+ dependent edema bilaterally. Again, the EP was concerned for a PE and ordered a repeat CTA scan of the chest. This study, similar to the first, was read as normal, and showed no evidence of PE. The patient was diagnosed again with “shortness of breath of unknown etiology” and discharged home. The patient’s obstetrician-gynecologist (Ob/Gyn) was not consulted; however, the patient was encouraged to follow up with him.

The next day, the patient presented to the same ED via emergency medical services, this time with seizures; she had no prior history of a seizure disorder. On presentation to the ED, she was noted to be postictal, with an elevated BP and tachycardic with an HR of 104 beats/minute. On examination, the lungs were clear to auscultation and the lower extremities exhibited 1+ pedal and pretibial edema. A urinalysis revealed proteinuria. The patient was given 4 g of magnesium sulfate intravenously (IV) and her Ob/Gyn was consulted.

The patient was admitted to the hospital with a diagnosis of eclampsia. She was given an IV drip of magnesium and labetalol for the high BP. Unfortunately, the patient apparently had suffered an anoxic brain injury from the previous seizures and died on hospital day 3.

The family sued the treating EPs and the hospital for failure to diagnose preeclampsia on two separate ED presentations. They noted the patient’s Ob/Gyn was never consulted; no action was taken to treat the hypertension; and no urinalysis was ordered on either visit. The EPs and hospital settled the case prior to trial for several million dollars.

Discussion

This is an incredibly sad case, and the EPs and hospital were right to settle and not go to trial. While PE was a reasonable diagnosis to consider in this patient on her first ED visit, it should not have been the only one in the differential diagnosis. The EP became anchored to this single diagnosis and refused to consider other alternative diagnoses—even after the CTA scan of the chest ruled out PE. Moreover, it appears the EP either never considered the significance of the elevated BP and dependent edema or just ignored these findings. To repeat essentially the same exact workup on the second visit does not make sense—one should “cast a wider net, not the same net.”

The diagnosis of “shortness of breath of unknown etiology” is similarly unacceptable. While this is a common and accepted diagnosis when it pertains to abdominal pain, the same is not true for dyspnea.

Preeclampsia is characterized by hypertension (BP >140/90 mm Hg) and proteinuria; associated symptoms include edema and hyperreflexia. Postpartum preeclampsia occurs infrequently and can develop up to 4 weeks after delivery.¹ In one 10-year retrospective case series, the incidence of preeclampsia in the postpartum period was 5.7%, and nearly 16% went on to develop eclampsia.² In a retrospective study of 22 postpartum preeclamptic patients, the median time to presentation was 5 days postpartum.¹ In a similar retrospective study of 152 patients, 90% of such patients presented within 7 days.³ The patient in this case initially presented on postpartum day 4.

 

Interestingly, in a study by Al-Safi et al,³ 63% of postpartum preeclamptic patients had no antecedent diagnosis of hypertensive disease during pregnancy. These findings are consistent with the findings of others that 33% to 69% of such patients show no evidence of preeclampsia in the ante- or peripartum period.

The clinical presentation of postpartum preeclampsia is similar to preeclampsia complicating pregnancy after gestation week 20. In the study by Al-Safi et al,³ headache was the most common presenting symptom (69%), followed by shortness of breath (30%), blurry vision (21%), nausea (12.5%), and epigastric abdominal pain (5%). Similarly, Yancey et al¹ found headache (82%) to be the most common presenting symptom in their series. Unfortunately, it is not known whether the patient in this case complained of headache or blurred vision as the published records note neither their presence nor absence.

The management of patients with preeclampsia includes IV magnesium to prevent seizures (ie, eclampsia) and BP control.¹ A bolus of 4 to 6 g IV magnesium sulfate over 15 to 30 minutes is recommended, followed by an infusion of 2 g/h IV. Historically, IV hydralazine has been used to manage preeclamptic patients with a BP greater than 160/110 mm Hg. More recently, however, IV labetalol has become popular.⁵ All such patients require admission to the hospital with Ob/Gyn involvement.

Missed Subdural Hematoma

A 59-year-old man presented to the ED with a chief complaint of headache, the onset of which he stated started gradually 2 days prior. He noted the headache was worse than normal but without associated nausea, vomiting, fever, chills, or change in vision. His past medical history was significant for a lower extremity deep vein thrombosis 3 months prior, for which he was taking warfarin.

The patient’s vital signs were all normal. The physical examination, including a thorough neurological examination, was also normal. The EP ordered a prothrombin time (PT), an international normalized ratio (INR), and a noncontrast CT scan of the head. The PT/INR results were therapeutic at 22 seconds and 2.3. The CT scan was interpreted by radiology services as normal. The patient’s headache was treated with IV prochlorperazine and diphenhydramine. After treatment, the patient reported feeling better and was discharged home with instructions to follow up with his primary care physician.

Over the next several months, the patient presented to the same ED on seven different occasions, each time with the chief complaint of headache. At each of these presentations, the history and physical examination were documented as unremarkable, with no history of trauma. The thoroughness, however, of the documentation varied considerably for each ED encounter. No head CT scan was ordered on the subsequent seven visits, and at each presentation, the patient was treated symptomatically and discharged home.

Two days after his eighth visit to the same hospital ED, the patient presented to a different ED, again with a chief complaint of headache. The EP at this ED ordered a noncontrast CT of the head, which demonstrated a left subdural hematoma. The patient was admitted to the hospital, given IV vitamin K and fresh frozen plasma, and underwent evacuation of the hematoma by neurosurgery. The patient’s hospital course was unremarkable, and he was discharged home without any focal weakness.

The patient, however, claimed that he suffered cognitive impairment as a result of the missed diagnosis. He sued treating EPs at the first ED as well as the hospital for failure to timely diagnose the subdural hematoma, stating that a CT scan should have been performed at each of his ED visits since he was on warfarin. The defense claimed that a CT scan was not warranted for each visit, and that the timing of when and how the brain bleed started was uncertain. At trial, a defense verdict was returned.

Discussion

It is well known that patients receiving warfarin are at an increased risk for intracranial hemorrhage (ICH) following blunt head trauma.¹ The recommendation is that all such patients have a noncontrast CT scan of the head to rule out intracranial bleeding. This is due to the fact that 60% of patients presenting with an immediate traumatic intracranial hemorrhage will have a normal mental status on examination; and 11% will have no history of loss of consciousness, a normal mental status examination, and no physical evidence of trauma above the clavicles.¹ In a study by Hart et al,² subdural hematoma accounted for 44% of all ICH in these types of patients.

More controversial is how to manage patients on warfarin who experience blunt head trauma and have a normal CT scan of the head. Because of the fear for delayed traumatic ICH, many clinicians recommend admitting such patients for neurological observation and repeat head CT scan the next morning.³ Additionally, some clinicians even recommend reversing the warfarin anticoagulation in such patients. ⁴ These recommendations, though, are based on expert consensus rather than on rigorous, prospective multicenter studies.¹  These strategies are also problematic, since such multiple repeat CT scans would not only be incredibly expensive but also would expose the patient to high doses of radiation to the brain. Moreover, the Centers for Medicare and Medicaid Services has now made CT brain scan imaging of patients presenting to the ED with complaint of nontraumatic headache a quality measure they follow. Their goal is to decrease the number of “unnecessary” head CT scans.

 

The patient in this case denied any history of trauma on the subsequent seven ED visits. Unfortunately, as pointed out, even minor trauma can result in ICH, and patients may not recall the occurrence of the event.

For patients on warfarin who present with headache, a very careful history must be taken—including inquiring about minor traumatic events. Even then, as has been shown, patients may have not experienced a loss of consciousness, have a normal mental status examination, and exhibit no external evidence of head trauma. The clinician is forced to use her or his own best judgment when evaluating such patients in the ED.

Interestingly, the risk of ICH secondary to blunt head trauma in patients on warfarin is increased if they are on concomitant aspirin therapy.² Similarly, the risk of ICH following head trauma in patients on clopidogrel is greater than for those patients taking warfarin,¹ and the risk of ICH in patients taking dabigatran is less than if taking warfarin.²

Missed Preeclampsia

A 24-year-old woman, gravida 1, para 1, aborta 0, presented to the ED complaining of a 1-day history of shortness of breath. Four days earlier, she had delivered a healthy baby boy via normal vaginal delivery and without complication. She denied chest pain, fever, or abdominal pain. She was otherwise in good health, stating that she was not taking any medications. She also denied smoking cigarettes.

On physical examination, the patient’s vital signs were remarkable for the following: heart rate (HR), 86 beats/minute; blood pressure (BP), 164/94 mm Hg; respiratory rate, 18 breaths/minute; temperature, 98.6^oF. Oxygen saturation was 96% on room air. The head, eye, ear, nose and throat examination was unremarkable. The lungs were clear to auscultation bilaterally, and HR and heart rhythm were normal. The abdomen was soft and nontender without guarding or rebound. The lower extremities were remarkable for 1+ pedal and pretibial edema bilaterally.

Since this patient was 4 days postpartum, the emergency physician (EP) was concerned for pulmonary embolism (PE). A complete blood count, basic metabolic profile, and a serum troponin T level were ordered. The electrocardiogram revealed normal sinus rhythm without evidence of strain or injury. The chest X-ray was interpreted by radiology services as normal. Given the concern for PE, computed tomography angiography (CTA) of the chest was ordered. All laboratory studies, including the troponin T level, were reported as normal. The CTA scan of the chest was interpreted by radiology services as normal and without evidence of PE. The patient was discharged home with a diagnosis of “shortness of breath of unknown etiology.”

The patient presented to the same ED 2 days later, again with the chief complaint of shortness of breath. On examination, her BP was noted to be elevated and she had 1+ dependent edema bilaterally. Again, the EP was concerned for a PE and ordered a repeat CTA scan of the chest. This study, similar to the first, was read as normal, and showed no evidence of PE. The patient was diagnosed again with “shortness of breath of unknown etiology” and discharged home. The patient’s obstetrician-gynecologist (Ob/Gyn) was not consulted; however, the patient was encouraged to follow up with him.

The next day, the patient presented to the same ED via emergency medical services, this time with seizures; she had no prior history of a seizure disorder. On presentation to the ED, she was noted to be postictal, with an elevated BP and tachycardic with an HR of 104 beats/minute. On examination, the lungs were clear to auscultation and the lower extremities exhibited 1+ pedal and pretibial edema. A urinalysis revealed proteinuria. The patient was given 4 g of magnesium sulfate intravenously (IV) and her Ob/Gyn was consulted.

The patient was admitted to the hospital with a diagnosis of eclampsia. She was given an IV drip of magnesium and labetalol for the high BP. Unfortunately, the patient apparently had suffered an anoxic brain injury from the previous seizures and died on hospital day 3.

The family sued the treating EPs and the hospital for failure to diagnose preeclampsia on two separate ED presentations. They noted the patient’s Ob/Gyn was never consulted; no action was taken to treat the hypertension; and no urinalysis was ordered on either visit. The EPs and hospital settled the case prior to trial for several million dollars.

Discussion

This is an incredibly sad case, and the EPs and hospital were right to settle and not go to trial. While PE was a reasonable diagnosis to consider in this patient on her first ED visit, it should not have been the only one in the differential diagnosis. The EP became anchored to this single diagnosis and refused to consider other alternative diagnoses—even after the CTA scan of the chest ruled out PE. Moreover, it appears the EP either never considered the significance of the elevated BP and dependent edema or just ignored these findings. To repeat essentially the same exact workup on the second visit does not make sense—one should “cast a wider net, not the same net.”

The diagnosis of “shortness of breath of unknown etiology” is similarly unacceptable. While this is a common and accepted diagnosis when it pertains to abdominal pain, the same is not true for dyspnea.

Preeclampsia is characterized by hypertension (BP >140/90 mm Hg) and proteinuria; associated symptoms include edema and hyperreflexia. Postpartum preeclampsia occurs infrequently and can develop up to 4 weeks after delivery.¹ In one 10-year retrospective case series, the incidence of preeclampsia in the postpartum period was 5.7%, and nearly 16% went on to develop eclampsia.² In a retrospective study of 22 postpartum preeclamptic patients, the median time to presentation was 5 days postpartum.¹ In a similar retrospective study of 152 patients, 90% of such patients presented within 7 days.³ The patient in this case initially presented on postpartum day 4.

 

Interestingly, in a study by Al-Safi et al,³ 63% of postpartum preeclamptic patients had no antecedent diagnosis of hypertensive disease during pregnancy. These findings are consistent with the findings of others that 33% to 69% of such patients show no evidence of preeclampsia in the ante- or peripartum period.

The clinical presentation of postpartum preeclampsia is similar to preeclampsia complicating pregnancy after gestation week 20. In the study by Al-Safi et al,³ headache was the most common presenting symptom (69%), followed by shortness of breath (30%), blurry vision (21%), nausea (12.5%), and epigastric abdominal pain (5%). Similarly, Yancey et al¹ found headache (82%) to be the most common presenting symptom in their series. Unfortunately, it is not known whether the patient in this case complained of headache or blurred vision as the published records note neither their presence nor absence.

The management of patients with preeclampsia includes IV magnesium to prevent seizures (ie, eclampsia) and BP control.¹ A bolus of 4 to 6 g IV magnesium sulfate over 15 to 30 minutes is recommended, followed by an infusion of 2 g/h IV. Historically, IV hydralazine has been used to manage preeclamptic patients with a BP greater than 160/110 mm Hg. More recently, however, IV labetalol has become popular.⁵ All such patients require admission to the hospital with Ob/Gyn involvement.

Missed Subdural Hematoma

A 59-year-old man presented to the ED with a chief complaint of headache, the onset of which he stated started gradually 2 days prior. He noted the headache was worse than normal but without associated nausea, vomiting, fever, chills, or change in vision. His past medical history was significant for a lower extremity deep vein thrombosis 3 months prior, for which he was taking warfarin.

The patient’s vital signs were all normal. The physical examination, including a thorough neurological examination, was also normal. The EP ordered a prothrombin time (PT), an international normalized ratio (INR), and a noncontrast CT scan of the head. The PT/INR results were therapeutic at 22 seconds and 2.3. The CT scan was interpreted by radiology services as normal. The patient’s headache was treated with IV prochlorperazine and diphenhydramine. After treatment, the patient reported feeling better and was discharged home with instructions to follow up with his primary care physician.

Over the next several months, the patient presented to the same ED on seven different occasions, each time with the chief complaint of headache. At each of these presentations, the history and physical examination were documented as unremarkable, with no history of trauma. The thoroughness, however, of the documentation varied considerably for each ED encounter. No head CT scan was ordered on the subsequent seven visits, and at each presentation, the patient was treated symptomatically and discharged home.

Two days after his eighth visit to the same hospital ED, the patient presented to a different ED, again with a chief complaint of headache. The EP at this ED ordered a noncontrast CT of the head, which demonstrated a left subdural hematoma. The patient was admitted to the hospital, given IV vitamin K and fresh frozen plasma, and underwent evacuation of the hematoma by neurosurgery. The patient’s hospital course was unremarkable, and he was discharged home without any focal weakness.

The patient, however, claimed that he suffered cognitive impairment as a result of the missed diagnosis. He sued treating EPs at the first ED as well as the hospital for failure to timely diagnose the subdural hematoma, stating that a CT scan should have been performed at each of his ED visits since he was on warfarin. The defense claimed that a CT scan was not warranted for each visit, and that the timing of when and how the brain bleed started was uncertain. At trial, a defense verdict was returned.

Discussion

It is well known that patients receiving warfarin are at an increased risk for intracranial hemorrhage (ICH) following blunt head trauma.¹ The recommendation is that all such patients have a noncontrast CT scan of the head to rule out intracranial bleeding. This is due to the fact that 60% of patients presenting with an immediate traumatic intracranial hemorrhage will have a normal mental status on examination; and 11% will have no history of loss of consciousness, a normal mental status examination, and no physical evidence of trauma above the clavicles.¹ In a study by Hart et al,² subdural hematoma accounted for 44% of all ICH in these types of patients.

More controversial is how to manage patients on warfarin who experience blunt head trauma and have a normal CT scan of the head. Because of the fear for delayed traumatic ICH, many clinicians recommend admitting such patients for neurological observation and repeat head CT scan the next morning.³ Additionally, some clinicians even recommend reversing the warfarin anticoagulation in such patients. ⁴ These recommendations, though, are based on expert consensus rather than on rigorous, prospective multicenter studies.¹  These strategies are also problematic, since such multiple repeat CT scans would not only be incredibly expensive but also would expose the patient to high doses of radiation to the brain. Moreover, the Centers for Medicare and Medicaid Services has now made CT brain scan imaging of patients presenting to the ED with complaint of nontraumatic headache a quality measure they follow. Their goal is to decrease the number of “unnecessary” head CT scans.

 

The patient in this case denied any history of trauma on the subsequent seven ED visits. Unfortunately, as pointed out, even minor trauma can result in ICH, and patients may not recall the occurrence of the event.

For patients on warfarin who present with headache, a very careful history must be taken—including inquiring about minor traumatic events. Even then, as has been shown, patients may have not experienced a loss of consciousness, have a normal mental status examination, and exhibit no external evidence of head trauma. The clinician is forced to use her or his own best judgment when evaluating such patients in the ED.

Interestingly, the risk of ICH secondary to blunt head trauma in patients on warfarin is increased if they are on concomitant aspirin therapy.² Similarly, the risk of ICH following head trauma in patients on clopidogrel is greater than for those patients taking warfarin,¹ and the risk of ICH in patients taking dabigatran is less than if taking warfarin.²

Missed Preeclampsia

A 24-year-old woman, gravida 1, para 1, aborta 0, presented to the ED complaining of a 1-day history of shortness of breath. Four days earlier, she had delivered a healthy baby boy via normal vaginal delivery and without complication. She denied chest pain, fever, or abdominal pain. She was otherwise in good health, stating that she was not taking any medications. She also denied smoking cigarettes.

On physical examination, the patient’s vital signs were remarkable for the following: heart rate (HR), 86 beats/minute; blood pressure (BP), 164/94 mm Hg; respiratory rate, 18 breaths/minute; temperature, 98.6^oF. Oxygen saturation was 96% on room air. The head, eye, ear, nose and throat examination was unremarkable. The lungs were clear to auscultation bilaterally, and HR and heart rhythm were normal. The abdomen was soft and nontender without guarding or rebound. The lower extremities were remarkable for 1+ pedal and pretibial edema bilaterally.

Since this patient was 4 days postpartum, the emergency physician (EP) was concerned for pulmonary embolism (PE). A complete blood count, basic metabolic profile, and a serum troponin T level were ordered. The electrocardiogram revealed normal sinus rhythm without evidence of strain or injury. The chest X-ray was interpreted by radiology services as normal. Given the concern for PE, computed tomography angiography (CTA) of the chest was ordered. All laboratory studies, including the troponin T level, were reported as normal. The CTA scan of the chest was interpreted by radiology services as normal and without evidence of PE. The patient was discharged home with a diagnosis of “shortness of breath of unknown etiology.”

The patient presented to the same ED 2 days later, again with the chief complaint of shortness of breath. On examination, her BP was noted to be elevated and she had 1+ dependent edema bilaterally. Again, the EP was concerned for a PE and ordered a repeat CTA scan of the chest. This study, similar to the first, was read as normal, and showed no evidence of PE. The patient was diagnosed again with “shortness of breath of unknown etiology” and discharged home. The patient’s obstetrician-gynecologist (Ob/Gyn) was not consulted; however, the patient was encouraged to follow up with him.

The next day, the patient presented to the same ED via emergency medical services, this time with seizures; she had no prior history of a seizure disorder. On presentation to the ED, she was noted to be postictal, with an elevated BP and tachycardic with an HR of 104 beats/minute. On examination, the lungs were clear to auscultation and the lower extremities exhibited 1+ pedal and pretibial edema. A urinalysis revealed proteinuria. The patient was given 4 g of magnesium sulfate intravenously (IV) and her Ob/Gyn was consulted.

The patient was admitted to the hospital with a diagnosis of eclampsia. She was given an IV drip of magnesium and labetalol for the high BP. Unfortunately, the patient apparently had suffered an anoxic brain injury from the previous seizures and died on hospital day 3.

The family sued the treating EPs and the hospital for failure to diagnose preeclampsia on two separate ED presentations. They noted the patient’s Ob/Gyn was never consulted; no action was taken to treat the hypertension; and no urinalysis was ordered on either visit. The EPs and hospital settled the case prior to trial for several million dollars.

Discussion

This is an incredibly sad case, and the EPs and hospital were right to settle and not go to trial. While PE was a reasonable diagnosis to consider in this patient on her first ED visit, it should not have been the only one in the differential diagnosis. The EP became anchored to this single diagnosis and refused to consider other alternative diagnoses—even after the CTA scan of the chest ruled out PE. Moreover, it appears the EP either never considered the significance of the elevated BP and dependent edema or just ignored these findings. To repeat essentially the same exact workup on the second visit does not make sense—one should “cast a wider net, not the same net.”

The diagnosis of “shortness of breath of unknown etiology” is similarly unacceptable. While this is a common and accepted diagnosis when it pertains to abdominal pain, the same is not true for dyspnea.

Preeclampsia is characterized by hypertension (BP >140/90 mm Hg) and proteinuria; associated symptoms include edema and hyperreflexia. Postpartum preeclampsia occurs infrequently and can develop up to 4 weeks after delivery.¹ In one 10-year retrospective case series, the incidence of preeclampsia in the postpartum period was 5.7%, and nearly 16% went on to develop eclampsia.² In a retrospective study of 22 postpartum preeclamptic patients, the median time to presentation was 5 days postpartum.¹ In a similar retrospective study of 152 patients, 90% of such patients presented within 7 days.³ The patient in this case initially presented on postpartum day 4.

 

Interestingly, in a study by Al-Safi et al,³ 63% of postpartum preeclamptic patients had no antecedent diagnosis of hypertensive disease during pregnancy. These findings are consistent with the findings of others that 33% to 69% of such patients show no evidence of preeclampsia in the ante- or peripartum period.

The clinical presentation of postpartum preeclampsia is similar to preeclampsia complicating pregnancy after gestation week 20. In the study by Al-Safi et al,³ headache was the most common presenting symptom (69%), followed by shortness of breath (30%), blurry vision (21%), nausea (12.5%), and epigastric abdominal pain (5%). Similarly, Yancey et al¹ found headache (82%) to be the most common presenting symptom in their series. Unfortunately, it is not known whether the patient in this case complained of headache or blurred vision as the published records note neither their presence nor absence.

The management of patients with preeclampsia includes IV magnesium to prevent seizures (ie, eclampsia) and BP control.¹ A bolus of 4 to 6 g IV magnesium sulfate over 15 to 30 minutes is recommended, followed by an infusion of 2 g/h IV. Historically, IV hydralazine has been used to manage preeclamptic patients with a BP greater than 160/110 mm Hg. More recently, however, IV labetalol has become popular.⁵ All such patients require admission to the hospital with Ob/Gyn involvement.

Missed Subdural Hematoma

A 59-year-old man presented to the ED with a chief complaint of headache, the onset of which he stated started gradually 2 days prior. He noted the headache was worse than normal but without associated nausea, vomiting, fever, chills, or change in vision. His past medical history was significant for a lower extremity deep vein thrombosis 3 months prior, for which he was taking warfarin.

The patient’s vital signs were all normal. The physical examination, including a thorough neurological examination, was also normal. The EP ordered a prothrombin time (PT), an international normalized ratio (INR), and a noncontrast CT scan of the head. The PT/INR results were therapeutic at 22 seconds and 2.3. The CT scan was interpreted by radiology services as normal. The patient’s headache was treated with IV prochlorperazine and diphenhydramine. After treatment, the patient reported feeling better and was discharged home with instructions to follow up with his primary care physician.

Over the next several months, the patient presented to the same ED on seven different occasions, each time with the chief complaint of headache. At each of these presentations, the history and physical examination were documented as unremarkable, with no history of trauma. The thoroughness, however, of the documentation varied considerably for each ED encounter. No head CT scan was ordered on the subsequent seven visits, and at each presentation, the patient was treated symptomatically and discharged home.

Two days after his eighth visit to the same hospital ED, the patient presented to a different ED, again with a chief complaint of headache. The EP at this ED ordered a noncontrast CT of the head, which demonstrated a left subdural hematoma. The patient was admitted to the hospital, given IV vitamin K and fresh frozen plasma, and underwent evacuation of the hematoma by neurosurgery. The patient’s hospital course was unremarkable, and he was discharged home without any focal weakness.

The patient, however, claimed that he suffered cognitive impairment as a result of the missed diagnosis. He sued treating EPs at the first ED as well as the hospital for failure to timely diagnose the subdural hematoma, stating that a CT scan should have been performed at each of his ED visits since he was on warfarin. The defense claimed that a CT scan was not warranted for each visit, and that the timing of when and how the brain bleed started was uncertain. At trial, a defense verdict was returned.

Discussion

It is well known that patients receiving warfarin are at an increased risk for intracranial hemorrhage (ICH) following blunt head trauma.¹ The recommendation is that all such patients have a noncontrast CT scan of the head to rule out intracranial bleeding. This is due to the fact that 60% of patients presenting with an immediate traumatic intracranial hemorrhage will have a normal mental status on examination; and 11% will have no history of loss of consciousness, a normal mental status examination, and no physical evidence of trauma above the clavicles.¹ In a study by Hart et al,² subdural hematoma accounted for 44% of all ICH in these types of patients.

More controversial is how to manage patients on warfarin who experience blunt head trauma and have a normal CT scan of the head. Because of the fear for delayed traumatic ICH, many clinicians recommend admitting such patients for neurological observation and repeat head CT scan the next morning.³ Additionally, some clinicians even recommend reversing the warfarin anticoagulation in such patients. ⁴ These recommendations, though, are based on expert consensus rather than on rigorous, prospective multicenter studies.¹  These strategies are also problematic, since such multiple repeat CT scans would not only be incredibly expensive but also would expose the patient to high doses of radiation to the brain. Moreover, the Centers for Medicare and Medicaid Services has now made CT brain scan imaging of patients presenting to the ED with complaint of nontraumatic headache a quality measure they follow. Their goal is to decrease the number of “unnecessary” head CT scans.

 

The patient in this case denied any history of trauma on the subsequent seven ED visits. Unfortunately, as pointed out, even minor trauma can result in ICH, and patients may not recall the occurrence of the event.

For patients on warfarin who present with headache, a very careful history must be taken—including inquiring about minor traumatic events. Even then, as has been shown, patients may have not experienced a loss of consciousness, have a normal mental status examination, and exhibit no external evidence of head trauma. The clinician is forced to use her or his own best judgment when evaluating such patients in the ED.

Interestingly, the risk of ICH secondary to blunt head trauma in patients on warfarin is increased if they are on concomitant aspirin therapy.² Similarly, the risk of ICH following head trauma in patients on clopidogrel is greater than for those patients taking warfarin,¹ and the risk of ICH in patients taking dabigatran is less than if taking warfarin.²