ACADEMIC EMERGENCY MEDICINE 2011; 18:78–85 © 2011 by the Society for Academic Emergency Medicine
Objectives: The patient with epidural hematoma and cerebral herniation has a good prognosis with immediate drainage, but a poor prognosis with delay to decompression. Such patients who present to nonneurosurgical hospitals are commonly transferred without drainage to the nearest neurosurgical center. This practice has never been demonstrated to be the safest approach to treating these patients. A significant minority of emergency physicians (EPs) have advised and taught bedside burr hole drainage or skull trephination before transfer for herniating patients. The objective of this study was to assess the effect of nonneurosurgeon drainage on neurologic outcome in patients with cerebral herniation from epidural hematoma.
Methods: A structured literature review was performed using EMBASE, the Cochrane Library, and the Emergency Medicine Abstracts database.
Results: No evidence meeting methodologic criteria was found describing outcomes in patients transferred without decompressive procedures. For patients receiving local drainage before transfer, 100% had favorable outcomes.
Conclusions: Although the total number of patients is small and the population highly selected, the natural history of cerebral herniation from epidural hematoma and the best available evidence suggests that herniating patients have improved outcomes with drainage procedures before transport.
Epidural hematoma is a collection of blood between the skull and the dura mater, usually caused by head trauma. The source is arterial in 36% of cases, venous in 32%, and presumably bone in most of the remainder.1 Those with arterial causes are more likely to need surgery, which is required in 44% to 90% of patients.2,3
Among those requiring surgery, it has been reported that 37% experience cerebral herniation syndromes, although this will vary with population.4 Blood accumulates at arterial pressure and crowds out cerebrospinal fluid and blood. Cerebrospinal fluid and blood represent 13% of the cranial volume, and once they are maximally displaced, the cranial compliance drops and intracranial pressure rises sharply.5 This causes herniation and ischemia.
Cerebral herniation is most commonly recognized clinically by deterioration in consciousness with unilateral or bilateral pupillary dilation.6 Other lateralizing signs include hemiparesis.4 Without rapid surgery, a herniation syndrome eventually causes permanent brain damage and death. Favorable outcomes are generally accepted as a Glasgow Outcome Scale (GOS) score of 4 (moderate disability) or 5 (good recovery; see Table 1).7,8
|1||Death||Death ascribed to brain damage|
|2||Persistent vegetative state||Unresponsive and speechless|
|3||Severe disability||Disabled and dependent|
|4||Moderate disability||Disabled and independent (e.g., dysphasia, hemiparesis, ataxia, intellectual, or personality changes causing social disruption).|
|5||Good recovery||Resumption of normal life (can have minor neurologic and psychological deficits)|
Patients with herniation due to epidural hematoma need immediate surgical decompression. In America, there are 5,759 hospitals and 3,200 neurosurgeons.9 Although our prehospital systems efficiently deliver most patients with severe head injuries to trauma centers, this issue can and does arise at any hospital. Extrapolating from epidemiologic data, the annual incidence of epidural hematoma is one case per 67,000 persons/year, and with herniation syndrome it is one case per 513,000 persons.10 Those emergency physicians (EPs) who work in areas without neurosurgical coverage can expect to see this presentation a handful of times in their careers. These estimates are similar to figures from other populations.11
The Brain Trauma Foundation and the Congress of Neurological Surgeons in 2006 recommended that surgical evacuation be performed “as soon as possible.”2 With prompt neurosurgical care, outcomes in patients with herniation syndromes can be surprisingly good. Some of the best outcomes ever reported occurred in a setting where neurosurgeons had a computed tomography (CT) scanner and an operating room in the emergency department (ED). Of 37 patients, 18 with a herniation syndrome, immediate surgery yielded a 95% rate of GOS 5.12 Favorable outcomes (GOS of 4 or 5) for patients with herniation syndrome from epidural hematoma in other studies range from 65% to 100%.4,7,13,14 The range is attributed to heterogeneity in the population.
In contrast, with delay to decompression, patients with herniation syndromes have increased rates of permanent disability and death.2 The critical time is not the time from injury, but the time from onset of symptomatic brain compression. A retrospective study addressing time from the onset of coma to operation in 32 patients found that mortality rose from 17% to 65% after 2 hours.15 Good recoveries (GOS 5) dropped after 2 hours from 67% to 13%.
In a prospective observational study with 74 herniating patients, favorable outcomes dropped progressively through time. With less than 90 minutes of herniation time, 78% had favorable outcomes, which dropped to 74% for 1.5 to 2.5 hours, to 69% for 2.5 to 3.5 hours, to 46% for 3.5 to 4.5 hours, to 33% for those with greater than 4.5 hours of herniation time.4 These studies included patients with other associated brain injuries, which dilutes the contribution of time to outcome. The small number of patients in each subgroup precludes using these data as a prognostic instrument.
Another retrospective study compared five patients with less than 70 minutes from the onset of anisocoria to decompression to five with times longer than 70 minutes. After 70 minutes, mortality rose from 0% to 100% and favorable outcomes dropped from 100% to 0%. This was a select population in that all patients developed their herniation syndromes after arrival.16
Therefore, the current literature varies somewhat, but shows a time-dependent worsening of outcome in patients with herniation syndromes. Despite this, a small but real proportion of patients can tolerate prolonged herniation. Most likely this represents variability in the rate and volume of bleeding. It is not possible in the ED to predict whether a given patient will be able to tolerate prolonged herniation. Based on available evidence, it is expected that with a several hour delay to decompression, the majority of patients will experience an unfavorable outcome. There is no time period that can be called a safe duration of herniation, likely because the herniation event is as heterogenous and unpredictable as the bleeding advancing it. The Congress of Neurological Surgeons notes that “every hour delay in surgery is associated with a progressively worse outcome.”2
As long as emergency medicine has been a specialty, a proportion of EPs have practiced and taught bedside skull trephination or burr hole drainage as a means of releasing intracranial pressure, restoring blood flow to the brainstem, and improving neurologic outcomes.17–22 It is hypothesized that this will improve rates of favorable neurologic outcome. The purpose of this review was to use an evidence-based format to identify anticipated neurologic outcomes in herniating patients with versus without drainage before transfer.
This was a structured literature review. The population is defined as those with cerebral herniation syndrome caused by epidural hematoma. A herniation syndrome is defined for the sake of simplicity as a deterioration in consciousness and pupil dilation.6 Contralateral motor deficits and Cushing’s reflex (bradycardia, hypertension, and diminished respiratory effort) are accepted as surrogate signs.22 The intervention is drainage of the hematoma by any nonneurosurgeon. This can be defined as drilling a hole in the skull overlying the hematoma using either a trephine or a perforator followed by burr bit or analogous approaches. Patients who are transferred without trephination are considered the control group, although actual controlled trials are usually not possible for conditions of extremely low frequency. The outcome of interest was neurologic function as measured by the GOS8 (see Table 1).
PubMed, EMBASE, the Cochrane Library, and the Emergency Medical Abstracts database were searched. The search strategy was “[epidural or extradural] hematoma outcome transfer” and “[epidural or extradural] hematoma outcome referred.” Additional articles were identified using reference lists and the related article function.
Articles were included if they identified outcomes in patients with epidural hematoma with herniation syndrome who presented first to nonneurosurgical hospitals. Articles were excluded if they did not attempt to include pupillary status and Glasgow Coma Scale (GCS) score at the first hospital. Articles were also excluded if their data did not isolate patients with herniation syndrome or if they used outcome in their selection criteria.
Outcomes for Patients Transferred Without Skull Trephination
For describing the control practice of transferring herniating patients without skull trephination, no articles were found meeting the methodologic criteria. Two articles were excluded from analysis because they lacked baseline pupillary data.23,24 One article was excluded because it was a case series of deaths from transfer, and thus outcome was part of the selection criteria.25 One of the articles was an article on EP skull trephination that mentioned the outcome in two patients that did not receive the procedure.26 The excluded articles are summarized in Table 2. For the sake of consistency, personal correspondence with one of the authors was used to differentiate moderate disability from combined moderate and severe disability (Poon WS, personal correspondence, December 18, 2009).
|First Author, Year||Type of Study||No. Patients With Herniation||Reason for Exclusion||Time From Symptoms to Surgery||Outcome GOS 4 or 5||Mortality|
|O’Sullivan, 1990||Retrospective||7||Lack baseline GCS/pupil data||NA||4/7||1/7|
|Poon, 1991||Prospective||12||Lack baseline GCS/pupil data||3.2 hours||2/12||8/12|
|Caird, 1999||Retrospective||4||Used outcome as selection criteria||NA||0/4||4/4|
|Smith, 2009||Retrospective||2||Lack baseline GCS/pupil data||150 min and 264 min||2/2||0/2|
Outcomes for Patients Transferred After Skull Trephination
For herniating patients with skull trephination before transfer, two articles met the methodologic criteria, one describing skull trephination by general surgeons, and one by EPs.26,27 Four articles were excluded for reasons described in Table 3.14,23,28,29
|First Author, Year||Type of Study||No. of Patients||Reason for Exclusion||Time From Symptoms to Decompression||Outcome GOS 4 or 5||Mortality|
|O’Sullivan, 1990||Retrospective||7||Lack of baseline GCS/pupil data||NA||6/7||0|
|K. Wester, 1999||Retrospective||11||Uses pupil data after operation, not before||NA||7/11||1/11|
|T. Wester, 1999||Retrospective||6||Lack of baseline GCS/pupil data||NA||4/6||1/6|
|Deverill, 2007||Retrospective||9||Lack of baseline GCS/pupil data||198 min||8/9||0|
Rinker et al.27
In 1990, a young man presented to the Bozeman Deaconess ED, in Bozeman, MT, with a left epidural hematoma, a right subdural hematoma, a GCS of 5 that deteriorated to 3, and a dilated right pupil. A local otolaryngologist who had learned burr hole craniotomy for his work in the international medical mission field intervened for the young man, who subsequently had a full functional recovery. After the event, the general surgeons at the hospital received workshop training by the receiving neurosurgeon in drilling burr holes. They reported results on five patients treated for epidural hematoma who met their criteria of having any of a GCS of 8 or less, pupillary dilation, or Cushing’s reflex.28 They performed the procedure on patients with subdural hematoma as well, although that is not the focus of our review. Four patients qualified for the procedure by pupillary changes and one by a GCS of 8. The surgeons did sometimes expand the burr hole; the technique was not specified, but is normally done by drilling an additional hole or using bone rongeur. All patients with epidural hematoma had favorable outcomes. All patients were transferred at once to the nearest neurosurgeon. Further clinical detail is provided in Table 4.
|Author, Patient||Diagnosis||Initial Exam||Exam Before Drainage||Pupil Exam||Exam After Drainage||ED Arrival to Decompression||CT to Decompression||Time to Craniotomy at Referral Center||GOS|
|Rinker, 1||EDH/SDH||5||3||Unilateral dilation||NA||63 min to OR||40 min to arrival in OR||NA||5|
|Rinker, 2||EDH||15||10||Unilateral dilation||NA||135 min to OR||50 min to arrival OR||NA||5|
|Rinker, 3||EDH||13||NA||Unilateral deviation||NA||90 min to OR||45 min to arrival OR||NA||5|
|Rinker, 4||EDH||8||8||Normal||NA||85 min to OR||45 min to arrival OR||NA||4|
|Rinker, 5||EDH||15||9||Unilateral dilation||NA||83 min to OR||35 min to arrival OR||NA||5|
|Rinker, 6||EDH||3||5||Unilateral dilation||NA||70 min to OR||35 min to arrival OR||NA||4|
|Smith, 1||EDH||Awake, talking, agitated. Soon intubated||NA||4 mm nonreactive||3 mm reactive||60 min||NA||180 min||5|
|Smith 2||EDH||alert, oriented||7 (then intubated)||dilated nonreactive||3mm, reactive||60 min||NA||180 min||4|
|Smith 3||EDH||15||7 (then intubated)||dilated, nonreactive||3mm reactive||NA||NA||NA||5|
|Smith 4||EDH||awake, alert, ambulatory||7 (then intubated)||dilated, nonreactive||reactive||60 min||NA||210||4|
|Smith 5||EDH||Reportedly returned to work||4, decerebrate||NA||NA||40 min||NA||120||5|
Smith et al.26
The second article reflects the approach taken by EPs trained at Hennepin County Medical Center in Minneapolis, MN.26 It is a retrospective review, mostly of patients with epidural hematoma referred to their local institution who had received prior skull trephination. All procedures were performed by EPs who had been trained in the procedure at Hennepin County Medical Center, either during residency training or through workshop courses. Patients were excluded from analysis if this procedure had been done in the operating room. Patients were included if they were initially lucid and then deteriorated, called “talk and deteriorate.” The authors identified five cases of epidural hematoma with herniation syndrome treated with skull trephination by EPs. Four had a dilated and unreactive pupil, and the fifth had a GCS of 4, with decerebrate posturing. EP intervention was skull trephination alone, without expansion by bone rongeur. They used suction to remove the blood. All four patients with dilated pupils had return of pupillary function after the procedure. In the remaining case, pupil function was not documented before or after the procedure. All patients had favorable outcomes, as described in Table 4. Brisk arterial bleeding was present in four of the five cases after the procedure. No patient had complications from the bleeding.
The literature on patients with epidural hematoma and herniation syndrome presenting to hospitals without neurosurgeons is sparse. No qualifying articles were found identifying outcomes in patients who are transferred without drainage. Among the excluded articles in Table 2, only eight of the 25 patients had favorable outcomes (eight of the 21 unselected patients).23–26 This is consistent with the natural history of herniation syndrome, where progressively worse outcomes are expected with time. At this time, we found no published studies showing that patients with cerebral herniation from epidural hematoma can be safely transferred without drainage.
It is important than any decision to transfer without drainage account for anticipated total herniation time or total time to craniotomy. Transfer time estimates should not be a projection based on distance, but ideally should be extrapolated from actual data. The time to craniotomy for a patient presenting with a herniation syndrome includes time spent at the first hospital, time spent in transport, and time spent in preparation at the receiving hospital. These times are rarely reported in full detail, but one recent representative article has done so. In a sample of 27 patients who needed “life-saving emergency craniotomy,” it counted medians of 7 minutes for the ambulance to respond to the patient, 33 minutes on scene, 11 minutes to arrive at the first ED, 45 minutes for the CT scan to be requested, 47 minutes for the CT scan to be performed, 18 minutes for the neurosurgeon to be called, 22 minutes for the patient to be accepted, 30 minutes for the ambulance to be contacted, 23 minutes for the ambulance to arrive, 15 minutes for the paramedics to load the patient, 45 minutes in transit (distance was between 19 and 85 kilometers), 0 minutes from arrival to being brought to the operating room (in 58% a room was reserved in advance), 28 minutes from arrival in the operating room to the start of surgery, and 15 minutes from the start of the case to decompression. The total median time to craniotomy was 5.4 hours.30 Only 14% of patients received craniotomy in less than 4 hours.
Although smaller case series might report times as fast as 2.9 hours, most larger articles do not.26 Metropolitan estimates for time to craniotomy in 39 patients were a median 5.25 hours.31 All hospitals were within 25 miles. A study of 46 patients in a more rural catchment area posted a median time of 8.1 hours to craniotomy.29 Seventy-nine percent of hospitals were within 200 km of the receiving hospital. Three of 20 herniating patients were able to get drainage within their own guideline of 2 hours. One article isolated transfer times for herniating patients (defined as head injury requiring neurosurgery with pupillary changes) and found that the total time delay was a median of 4.27 hours for patients with such changes.32 They did not include time from arrival at the receiving hospital to surgery, which adds another 45, 84, or 134 minutes.29,31,33 Adding the minimum estimate for the additional time would take the total time to craniotomy above 5 hours. It is important that transfer policies reflect what actually happens, not what is desired should happen, and at this time, it appears that patients in need of emergency neurosurgery commonly receive transfer delays of over 5 hours.
One study that started the clock at the patient’s time of deterioration did show a mean transfer time of 3.2 hours, although this did not include time from arrival to operative decompression.18 Eight of their 12 patients died. No other study has ever shown that the majority of patients can receive time to craniotomy less than four hours.
Adequacy of a Single Burr or Trephination Hole
Evidence has accumulated that skull trephination alone can improve neurologic outcomes. This appears to be true even when the hematoma is not completely evacuated. In a study from Norway, only three of 11 patients had technically “adequate” procedures with full and persistent resolution of the hematoma.14 Despite high rates of residual hematoma, all seven patients with burr hole drainage had favorable outcomes. The four patients with unfavorable outcomes had full craniotomy. It is not probable with a full craniotomy that the unfavorable outcomes resulted from incomplete release of pressure. Individual unfavorable outcomes likely have more to do with the magnitude of the initial injury and the time to initial pressure release. Therefore, even with incomplete hematoma evacuation, burr hole drainage by nonneurosurgeons appears to be effective in improving outcomes. Presumably this is through release of enough pressure to restore blood flow to the brainstem and arrest the herniation event. Nonetheless, as much hematoma as possible should be removed through suction.
A neurosurgeon in Taiwan evaluated the sufficiency of burr hole drainage through a suction catheter in 13 patients with epidural hematoma.34 The purpose of the study was to establish “an alternative method for rapid and efficient clot evacuation ... thereby allowing more patients to be saved within shorter time periods.” It was successful in 11 cases, whereas in two cases the catheter clogged and a full craniotomy was employed. Five of the procedures were performed in the ED due to rapid deterioration of the patient. All 13 patients apparently made full recoveries. None had bleeding complications, and two patients required a second trephine hole. The author concludes, “the procedure would be appropriate for rapid management of a traumatic brain injury patient in rural or underdeveloped areas with no alternative definitive neurological services ....”34
After reviewing the evidence for improved outcomes with burr hole drainage, the Royal Australasian College of Surgeons recommended that country practitioners perform burr hole drainage for herniating epidural hematoma if transfer was expected to take longer than 2 hours.35 Where this policy was followed in that region, eight of nine patients so treated had favorable outcomes.29 All available evidence at this time (both of the excluded articles in Table 3 and of the included articles in Table 4, as well as various case reports) support the efficacy of single hole drainage in protecting the patient from brain damage.36,37
To keep the procedure simple, a single hole with suction should probably be the whole procedure in most cases, although expansion with a bone rongeur or a second hole could be considered if the physician finds it necessary. In general, the goal of the procedure is not to remove all of the hematoma, but to release the excessively high pressure that is causing cerebral herniation. This is analogous to other situations, such as pericardial tamponade, where a steepened compliance curve means that a small decrease in volume yields a large decrease in pressure. For the herniating patient, improving the perfusion of the brain yields future quality of life improvements.
Safety of Skull Trephination
The main risks of drainage by trephine or burr include bleeding, plunging, and infection. By definition the patient already has uncontrolled bleeding, with the brain in a sense compressing the hematoma. A single burr hole or trephination by itself generally does not allow hemorrhage control. Instead, the patient is allowed to bleed. For the five cases reported by Smith et al.,26 four of the five patients had brisk or arterial bleeding, yet no patient had complications from the bleeding. Nonetheless, until more is known, the patient should likely be sent with group O blood to be available for transfusion along the way. Patients with hemostatic deficiencies should receive the appropriate reversal agents and hemostatic resuscitation such as platelets, fresh-frozen plasma, vitamin K, etc.
Additional bleeding risks would include laceration of venous sinuses. The frontotemporal region, where most of the significant hematomas predominate, is relatively free of venous sinuses and is a safe place to perform skull trephination. CT localization is used to confirm the site, and if the hematoma is located away from traditional drilling sites, the EP should discuss the site with the neurosurgeon. Traditional temporal diagnostic burr holes were done approximately two fingerbreadths above the zygoma and two fingerbreadths anterior to the auditory canal.20
The second main risk for this procedure is plunging. Plunging has been reported at least once in their career by as many as 66% of neurosurgeons.38 However, 95% of these were with an unclutched Hudson brace with a perforator drill bit. Most nonneurosurgeons would tend to use either clutched drills or a trephine, which are less likely to cause plunging. With proper training, plunging should generally be very rare. The subjacent hematoma acts as an additional buffer against plunging if the site is chosen well. No articles in this review were found reporting any plunging by nonneurosurgeons.
The risk of infection from penetration of the cranial cavity is low and acceptable, even when done outside of the operating room. One group of neurosurgeons performed trephination for herniating patients in the CT scanner and reported no infection after 26 cases.39 When intensive care physicians place intracranial pressure monitors, the rate of meningitis ranges from 0.5% to 0.8%.40,41 Nurse practitioners and physician assistants safely perform this procedure as well.42 The rate of meningitis after craniotomy in general is reported as 0.8% to 1.5%.43 To summarize, the procedure is safe with an acceptable risk profile with regard to bleeding, plunging, and infection. The benefits of improving neurologic outcomes appear to far outweigh the risks of the procedure.
It is not desirable that patients with epidural hematoma present to nonneurosurgical hospitals, but inevitably this will occur. When the patient has signs of cerebral herniation the condition is critical, and the transfer process means that the median total time to craniotomy in most cases is more than 4 hours. At this time there is no evidence that patients with epidural hematoma and signs of herniation can be safely transferred without drainage. Ultimately the question of how to treat these patients must be decided based on outcomes. Drainage by nonneurosurgeons before transfer has thus far yielded favorable outcomes. The articles in this review showed 100% of patients with cerebral herniation had favorable outcomes, although the excluded articles present a more realistic 64% to 100%. Neurosurgeons’ rates of favorable outcome in herniating patients range from 65% to 100%.4,7,13,14 The populations and procedures are not identical, and the numbers are low, so this is not to suggest that EP drainage will work in all patients or that it will necessarily be as effective as neurosurgical drainage. However, if there is any selection bias in these articles, it might have tended to favor the procedure being done in a sicker population (patients critical enough to deem the decision to drill unimpeachable).
The available data are limited at this time, so recommendations should not overstate the argument for this procedure. Judgment is required in each individual case as to benefits and risks of drilling versus not drilling. Unanswered questions include the proper hole size, whether suction alone is sufficient to remove most of the clot, and whether some patients will experience hypovolemia from blood loss. Additionally, this article has included physicians from different specialties, and more than half of the procedures were performed by general surgeons. The skills for trephination alone probably do not differ significantly between specialties, but more data are needed to report outcomes for the procedure as performed by EPs. Further studies should address these questions, as well as optimal training and skill retention.
Obstacles to Implementation
It is not common that a patient will need this procedure to be performed by an EP. However, when the need arises, this review suggests that EP bedside skull trephination has the potential to greatly improve neurologic outcome. For this procedure to be performed, certain obstacles need to be addressed.
The first is a concern that the receiving neurosurgeon might be skeptical that an EP can perform the procedure. The EP must do his or her part by being skilled and prepared for the procedure and be able to relate the bedside status of the patient and why there is concern for a grave prognosis without drainage. In their articles, neurosurgeons frequently comment on the need for drainage to be done by nonneurosurgeons in emergency conditions.23,25,34,44,45 In the case reports written by EPs, it was often the neurosurgeon who advised the procedure be performed.37,46 The media has reported several cases of EPs saving patients’ lives with household drills, but these were all decisions made in counsel with neurosurgeons.47–49 This procedure represents an opportunity for collaboration between EPs and neurosurgeons in a lifesaving endeavor that can leave practitioners of both specialties feeling satisfied about the decision.
The second obstacle is training. The logistic challenges in training EPs in skull trephination are actually quite small. The procedure does not require advanced skills. On a technical level, it is simpler than many of the other procedures performed by EPs. For the articles included in this review, formal instruction was completed in a single day.26,27 Even without training, it has been learned in real time, by reading a faxed report from the neurosurgeon.37 It has also been performed by nonphysicians, such as in the case where this was done by a dentist who used his dental drill.46 As a specialty, however, we must ask that the procedure be learned well in advance. In emergency medicine we are called upon to perform rare but lifesaving procedures such as cricothyrotomy and pericardiocentesis. These skills are taught to residents in procedure laboratories and to practicing physicians at conferences. Burr hole drainage and skull trephination should be taught in the same way. Every EP should be expected to be able to perform this procedure effectively.
Resources for instruction include the World Health Organization’s free online textbook Surgical Care at the District Hospital, which has a chapter called “Cranial Burr Holes.”50 The procedure is also reviewed in several articles and emergency medicine procedure books.19–21,26,36,37,44,45,51 Telemedicine consultation for this procedure has now been reported and will likely help increase the comfort level of EPs by bringing the neurosurgeon’s experience and guidance to the bedside.52 In the future, perhaps neurosurgeons will be able to perform the procedure remotely by robotics.
The final obstacle is a perception of medicolegal risk. Skull trephination at this time is not part of the core curriculum of emergency medicine residency, and therefore an EP might feel medicolegally safe deferring this procedure to the tertiary care center. Regardless of whether that argument has legal validity, it is probably not the moral or ethical way to approach this problem. No patient should ever be denied the opportunity for health when it is available. EP skull trephination is a technique that has been shown to dramatically improve outcomes in selected patients. With good communication, including informed consent when possible, disclosure of our limitations as nonneurosurgeons, and discussion with the accepting neurosurgeon, the medicolegal risk of this procedure is probably trivial. Most families would likely appreciate that their loved one at least had a chance.
The primary limitation of this review is the scarcity of patients. The included articles reflect outcomes in a total of 11 patients who received nonneurosurgeon trephination or burr hole drilling. For patients not receiving drainage before transfer, no cases were reported that met methodologic criteria. Even when counting the excluded articles, the total number of patients involved was 25 who did not receive drainage and 43 who did. These articles themselves are retrospective and subject to publication bias.
The patient population itself is heterogeneous, with a rural population described in the article by Rinker et al. and a metropolitan population described in the article by Smith et al. The training and skill set for the general surgeons and emergency physicians is potentially different as well. Finally, the procedures performed were heterogenous, comprising both burr hole drainage and skull trephination.
Based on this literature review, patients with local nonneurosurgeon drainage of epidural hematoma prior to transfer to neurosurgical care appear to have consistently favorable outcomes, and therefore the data at this time are sufficient to support its practice. It can reverse or temporize a herniation syndrome with an acceptable risk profile. Because it has the potential to dramatically affect the lives of the patients who qualify for it, it should be taught as a life-saving intervention that is part of the potential skill set of the EP.