neurosurgical intervention: a retrospective study of the...
TRANSCRIPT
Neurosurgical Intervention: A retrospective Study of the Morbidity and Mortality associated with
Craniotomies and Craniectomies for treatment of Severe Traumatic Brain Injury.
John Ogunlade, D.O., Fanglong Dong PhD, Dan Miulli D.O.
Abstract
Objective:
To identify the relative rate of morbidity and mortality after Immediate Surgical decompression
(via craniotomy or craniectomy) in a youthful population who have suffered a severe traumatic brain
injury, in order to better predict post-surgical outcomes.
Methods:
This single institution retrospective study used inclusion and exclusion criteria to identify sample
subjects from 10,779 consecutive patients from the ARMC neurosurgery database. A total of 50 patients
were included in this study and a meticulous review of their medical records identified multiple
independent factors for comparison of patient's primary outcome, Glasgow outcome score at time of
hospital discharge, and secondary outcome, preservation of life. For this study, favorable outcome is a
GOS of 4 or 5, Morbidity is defined as GOS of 2 or 3, and Mortality is defined as Death or GOS of 1.
Results:
A total of 50 patients were included in the final analysis. Half (48%) of them had a Craniotomy
and the other half (52%) had a Craniectomy. Patients who were taken for a Craniotomy were more likely
to have a presenting GCS between 6 and 8 (70.8% vs 38.5%, p=0.0218), had higher chance of post-
operative survival (100% vs 76.9%, p=0.0121), and had a shorter mean hospital length of stay (13 days vs
34 days, p=0.2814, statistically significant at p <0.01). Additionally, patients who received a Craniotomy
had a smaller rate of complications (4.2% vs 19.2%, p=0.1015), though the difference was not
statistically significant.
Conclusion:
Youthful patients who suffer from severe traumatic brain injury have a significant chance of
meaningful survival with emergent surgery. Based on the mechanism of injury, the presenting GCS and
the surgery performed, the probability of significant disability and/or mortality can be inferred using the
Severe TBI Disability Score. Patients who undergo craniotomy as opposed to craniectomy have better
outcomes as a result of the severity of injury sustained rather than surgery performed. The long term
outcome of patients who require craniectomy is unpredictable, however it may be the patients best
chance at survival. The Severe TBI Disability score is a tool that could be used to help explain to families
the relative probability of 30 day mortality and long term disability. It is not intended to direct treatment
or surgical management of TBI Ultimately the decision to provide any kind of treatment to patients with
a severe TBI is the responsibility of the attending Neurosurgeon.
Introduction:
Traumatic brain injury is a life changing event that affects the patients, their families and the
community at large. In 2013, an estimated that 2.8 million people suffered from Traumatic Brain Injuries
in the United States alone[1]. Of that, 282 thousand people were hospitalized, and aproximately 50
thousand people died from TBI [1]. In the setting of trauma, Neurosurgical intervention is considered a
lifesaving and heroic measure; however the quality of life after surgery is unpredictable and is riddled
with long hospital stays, persistent vegetative state, and death despite aggressive treatment.
A landmark randomized study concluded that patients who underwent bifrontotemporoparietal
decompressive craniectomies to treat refractory elevated intracranial pressures had worse scores on
their Extended Glasgow Outcome Scale, and no statistically significant difference in rate of death at 6
months. This was in comparison to patients treated with standard care, defined as treatment by
external ventricular drainage of cerebral spinal fluid, optimized sedation, normalizing arterial carbon
dioxide pressure, and the use of mannitol, hypertonic saline, and neuromuscular blockade. The study
randomized its patients into treatment groups within a 72 hour period with surgical decompression
occurring after the randomization. In this study, the median time to randomization and surgical
decompression was 35 hours, as opposed to immediately. The extended period of refractory elevated
intracranial pressures in both groups may explain similar rates of mortality seen at 6 months. In addition
patients were taken back to surgery for replacement of their bone flap at 2 to 3 months from
decompression, then subsequently evaluated using the Extended Glasgow Outcome Scale at 6 months.
It is possible that the 6 month evaluation is obscured as a result of the compounding effects of the
second surgery in addition to the first, or premature cranioplasty.
Another study compared Craniotomy to Craniectomy in the treatment of Acute Subdural
Hematoma and concluded that patients who undergo craniectomy are younger, have significantly longer
hospitalizations, and are more likely to be discharged to a skilled nursing or rehabilitation facility or to
die in the hospital. This study performed a propensity match based on a number of demographics
(including age, sex, race, insurance status, (coverage or no coverage), length of stay, in-hospital
mortality, and discharge disposition), however it failed to match presenting Glasgow Coma Scales, or
other compounding injuries. It is likely that the craniotomy group had better results in comparison to
the Craniectomy group because the craniectomy group likely suffered a more severe injury, had a bigger
subdural hematoma with larger midline shift, and or a significant amount of brain swelling seen
intraoperatively.
A more recent study suggest that the preoperative cerebral midline shift, initial GCS and
presence of intraventicular hemorrhage were predictors of whether the bone flap should be kept off
after the evacuation of subdural hematoma[9]. Nguyen et al reported that patients in their study who
had larger presenting midline shift exhibited a negative change in brain volume, which was
representative of observed cerebral edema, thus suggesting that replacement of the bone flap is
reasonable in patients with larger midline. This is misleading as the patient’s midline shift varies based
on patient’s brain compliance, age related cerebral atrophy, or mechanism of subdural, e.g. severe
trauma versus spontaneous SDH related to anticoagulant use.
This current study is an attempt to better understand the morbidity and mortality associated
with severe traumatic brain injury that is treated with immediate neurosurgical intervention in the form
of surgical decompression (Craniotomy vs. Craniectomy). The current study differs in that, first; it is a
retrospective study that looks at Glasgow Outcome Scores at time of discharge as the primary outcome
measure after immediate surgical decompression. Second, the patients of this study represent a
youthful population where age is not the limiting prognostic factor. Third, the initial GCS exam
performed by the Neurosurgeon is not masked by the effects of illicit drug use, and last, post surgical
treatment is not complicated with the management of pre-existing conditions, medical problems or
withdraws from substance abuse.
Methods:
The current study is a single institution retrospective chart review using inclusion and exclusion
criteria to identify sample subjects from a database of 10779 consecutive patients. After institutional
review board approval was obtained, the ARMC neurosurgery patient database from 2005 to 2015 was
used to identify all patients from age 18 to 30 who were seen by the Neurosurgery Service at ARMC
(543). The Medical records of each patient were reviewed to identify any possible exclusion criteria.
After the study population is identified (n=50), an intense and thorough chart review of each patient is
completed and the following independent variables are collected for statistical analysis; Age, Gender,
Injury sustained, Surgery performed, Presenting GCS, Intubated on presentation, Last GCS documented,
Length of Hospital stay, time till last follow up, Time to Cranioplasty (for patients who had a
craniectomy), surgical complications, Glasgow Outcome scale, and overall out come. In this study,
favorable outcome is defined as GOS of 4 or 5, Morbidity is defined as GOS of 2 or 3, and Mortality is
defined as Death or GOS of 1.
Inclusion criteria:
1. Patients between the ages of 18 to 30
2. Suffered a severe traumatic brain injury, defined as initial presentation to ED as GCS of 8 or less.
3. Required Neurosurgical intervention in the form of a Craniotomy or Craniectomy and was taken to
surgery emergently on presentation.
Exclusion Criteria:
1. History of illicit drug use, Positive Urine Drug screen for Methamphetamine, Cocaine, Cannabis, PCP
or ETOH level > 0.08
2. History of pre-existing Medical disease for which the patient is being treated (e.g. Brain tumor, AVM,
previous brain surgery, HTN, DM, coagulopathy, renal disease).
3. Concurrent Traumatic injury requiring other specialty surgery (e.g. open abdomen, Long bone
fracture, unstable spinal fracture.)
Methods in Detail:
1. The Neurosurgery patient Database from June 2005 to April 2015 (10779 patients) is interrogated for
patients who meet the age inclusion criteria, creating the survey list (543 patients).
2. Each patient’s medical record is briefly reviewed to included patients who suffered traumatic brain
injury which required immediate neurosurgical intervention and exclude patients based on the exclusion
criteria. The patients who do meet the exclusion criteria are removed from the survey list creating Short
List (105 patients).
3. Comprehensive chart review is completed on each patient on the Short list and the following
information is identified: Age, Gender, injury sustained, Surgery performed, Presenting GCS, Intubated
on presentation, Last GCS documented, Length of Hospital stay, time till last follow up, Time to
Cranioplasty (for patients who had a craniectomy), surgical complications, GCS Outcome Scale, and
overall out come.
4. Any patient who’s presenting GCS was not 8 or less was removed from the study creating the final
patient list (50 Patients).
5. Data is analyzed for statistically significant differences.
Results:
A total of 50 patients were included in the final analysis. Majority (87%) of them were males,
had an average age of 22.38 ± 3.12 years, had Craniotomy (48%), experienced closed head injury (80%),
and 54% had GCS 6 to 8 upon on admission into the hospital. Majority of patients survived the surgery
(88%), returned to a normal GCS value (n=79, 75.2%), had no complications (88%), and had a median of
13.5 days for hospital stay. A more detailed description of variables was presented in Table 1.
Crosstab analyses were conducted to identify factors by Craniectomy or Craniotomy. Table 2
presented the analysis results. Patients who received Craniotomy were more likely to have a presenting
GCS between 6 and 8 (70.8% vs 38.5%, p=0.0218), had higher post-operative chance of survival (100% vs
76.9%, p=0.0121), and had a longer average length of stay (13 days vs 34 days, p=0.2814, statistically
significant at p <0.01). Additionally, Patients who received Craniotomy were associated with less
complications (4.2% vs 19.2%, p=0.1015), though the difference was not statistically significant.
Discussion:
A youthful patient population that is free of preexisting medical conditions, and whose
neurosurgical evaluation, intervention and hospital course is not obscured by the effects of illicit drugs
or alcohol abuse serves as the best possible control group to evaluate the effectiveness of
Decompression after Severe TBI. The Kennard principle suggest that the developing brain in the
youthful population can recover and reorganize after injury[7]. Furthermore, the younger brain, in
contrast to the elderly brain, is less likely to develop progressive cognitive decline [6], and the ongoing
development seen in youth may in actuality promote recovery in the presences of structural brain
damage, which is unlikely to occur in a fully matured adult brain, Berger et al[8]. The reported rates of
morbidity and mortality in TBI patients remain inconsistent and unpredictable likely due to the large age
distribution in which TBI occurs, which is then further convoluted by the complications that arise from
the compounding effects of secondary brain injury magnified by preexisting medical diseases.
The current study strategically focuses on a healthy and youthful population of patients who
suffered a severe traumatic brain injury and was taken emergently for surgical decompression. All
patients who underwent craniotomy survived, with the majority (87.5%) of these patients having a
favorable outcome, a Glasgow Outcome Score of 5 (62.5%) and 4 (25%), representing low to moderate
disability at time of discharge (average length of stay was 11 days). In the Craniectomy group, 6 patients
(23.1%) died despite emergent surgical decompression. All 6 of these patients presented with a GCS
between 3 to 5, which represents a 26% rate of mortality associated with a presenting GCS of 3 to 5.
There is a 37.5% likelihood of death when a craniectomy is performed on a patient presenting with a
GCS of 3 to 5. Independent of other factors, Decompressive Craniectomy was associated with only
23.1% mortality rate.
No mortality was seen in the craniotomy group, however this does not suggest that a
craniotomy is a safer surgery. Mortality in Severe TBI is an inherent result of the force delivered from
mechanism of injury, as all patients who died after emergent surgical decompression suffered a Closed
Head injury. The results show that there is a 15% chance of mortality associated with Closed head
Injuries, and the majority of Closed Head Injury patients (62.5%) required a craniectomy due to Cerebral
swelling seen intraoperatively.
Morbidity in this study is defined as patients who survived with significant disability
representing a Glasgow Outcome Score of 3 and 2. In the craniotomy group the total rate of morbidity
was 12.5%. On presentation, the GCS of these 3 patients were 3T, 4T and 6T, and their GCS at discharge
was 10T, 6T and 15 respectively. At the time of discharge the patient that was a GCS of 15 still needed
24 hour care and had considerable motor dysfunction, dysarthria and dysphagia which required a
Gastrostomy Tube for feedings. Within the 2 year period that the patient followed up, a
Ventriculoperitoneal shunt was placed after he developed post traumatic communicating
hydrocephalus, however he was cleared for normal activity at his last follow up visit.
Of the remaining 20 patients who survived after a decompressive craniectomy, only 6 (23.1%)
had a favorable Glasgow Outcome Scale score of 5 (1 patient) or 4 (5 patients) at time of hospital
discharge. Morbidity was significantly higher at 53.9%, as a total of 14 patients had a GOS of 2 or 3 at
time of hospital discharge. Of these 14 patients, 7 were lost to follow up. Of the 7 remaining patients, 5
of 7 required ventriculoperitoneal shunt placement for post traumatic communicating hydrocephalus
and 3 of 7 developed seizure disorder that required continued follow up with neurology. Overall the
Glasgow outcome score for these patients remained unchanged at last neurosurgical follow up, which
averaged 1.7 years.
All but 1 patient (96.1%) in the craniectomy group suffered a closed head injury, which also
supports that closed head injuries are more commonly related to cerebral swelling seen
intraoperatively, thus suggesting a higher force impact and a larger magnitude of brain injury, resulting
in an unfavorable outcome such as severe disability, vegetative state and death.
Complications were seen primarily in the craniectomy group. A total of 6 surgical complications
were identified with 5 of 6 occurring in the craniectomy group (see table 2). Of the patients who
survived, 13 had a successful cranioplasty, with the mean time to cranioplasty being 12 months. A total
of 7 patients were lost to follow up and did not have a cranioplasty at ARMC.
There were 2 patients who presented with penetrating gunshot wounds to the head, both
presented as a GCS of 7T (intubated), and both underwent an emergent Craniotomy. The Patient who
had the shorter length of stay of 4 days had a Glasgow Outcome scale of 5, while the other patient
whose length of stay was 12 days had a GOS of 4. Both patients returned to a GCS of 15 prior to
discharge. These patients were fortunate to have reached the hospital alive and intubated, as many
guns shot wound victims do not; and if they do, death and severe disability is often presumed. The
decision for emergent surgical intervention should be evaluated on a case by case bases, these 2 cases
are merely examples of favorable outcomes in what may have been presumed fatal.
Though the study started with a database slightly over ten thousand patients, only a total of 50
patients remained after the inclusion and exclusion criteria were applied. In most studies, such a small
sample size is not representative of a much larger population, and this same principle holds for this
current study. However the information provided by this intense retrospective review of a specific
population, who should theoretically have the best chance of surviving a severe TBI, gives the reader a
relative account of Morbidity and Mortality associated with the implication of being able to complete a
craniotomy versus reverting to craniectomy and the intrinsic nature of severe TBI independent of the
negative effects associated with advanced age, substance abuse or preexisting conditions.
Conclusion:
Young patients who suffer from severe traumatic brain injuries have a significant chance of
survival and great potential for functional recovery with emergent surgery. Though it is impossible to
predetermine the magnitude of the brain injury endured, based on the mechanism of injury, the
presenting GCS and the surgery performed, the Severe TBI Disability Score infers the probability of
significant disability and/or mortality from the results presented in this study. The degree of cerebral
swelling seen intraoperatively determines whether the bone flap can be replaced once the surgery is
completed, and is thus indicative of the degree of injury sustained by the brain. Therefore, the overall
better outcome seen in the craniotomy group in this study is a result of a less severe injury rather than
the type of surgery performed. Each patient should be evaluated independently of previous outcomes,
and the decision to perform a craniotomy or craniectomy should be made in the operating room on a
case by case basis. The results presented in this study may be helpful in outlining the long term outlook
for patients and their families after heroic surgical measures have been performed.
The Severe TBI Disability score is a tool that should be used to help explain to families the
relative probability of 30 day mortality and long term disability. It is not intended to direct treatment or
surgical management of TBI and does not apply to patients who do not meet the inclusion or exclusion
criteria. The outcome of patients who have Penetrating brain injuries varies based on many presenting
factors and therefore should be assessed on an individual case by case basis. Ultimately the decision to
provide any kind of treatment to patients with a severe TBI is the responsibility of the attending
Neurosurgeon.
Severe TBI Disability Score
Presenting GCS 3-5 1
6-8 0
Surgery Performed Craniectomy 1
Craniotomy 0
Injury Sustained Closed Head Injury 2
Depressed Skull Fracture 1
30 day percent chance of severe disability (GOS < 4)
4 Points 87% *
3 Points 64%
2 Points 6%
1 Point <1%
*- includes 40% chance of Mortality and 47% chance of severe disability
if patient survives 30 days
Table 1: Patient demographic information (N=50)
Frequency (N=50) Percent
Sex
Female 7 14%
Male 43 86%
Craniotomy_vs_Craniectomy
Craniectomy 26 52%
Craniotomy 24 48%
Mechanism of injury
CLOSED HEAD INJURY 40 80%
DEPRESSED SKULL FRACTURE 8 16%
GSW 2 4%
Complications
CNS infection (RTO) 3 6%
CSF Leak (RTO) 1 2%
None 44 88%
Rebleed (RTO) 2 4%
Mortality
Death 6 12%
Survived 44 88%
GCS_before_surgery_2_group
GCS 3 to 5 23 46%
GCS 6 to 8 27 54%
Glasgow Outcome Scale
1 6 12%
2 5 10%
3 12 24%
4 11 22%
5 16 32%
AGE 22.38 ± 3.12
Length_of_Stay__Day (mean) 24
Table 2: Crosstab analysis of factors associated with the two surgical procedures
Craniectomy Craniotomy P-value
GCS before surgery 0.0218
GCS 3 to 5 16 (61.5%) 7 (29.2%)
GCS 6 to 8 10 (38.5%) 17 (70.8%)
Glasgow Outcome Scale <.0001
1 6 (23.1%) 0 (0%)
2 4 (15.4%) 1 (4.2%)
3 10 (38.5%) 2 (8.3%)
4 5 (19.2%) 6 (25%)
5 1 (3.9%) 15 (62.5%)
Mortality 0.0121
Death 6 (23.1%) 0 (0%)
Survived 20 (76.9%) 24 (100%)
Complications 0.1015
CNS infection (RTO) 3 (11.5%) 0 (0%)
CSF Leak (RTO) 0 (0%) 1 (4.2%)
None 21 (80.8%) 23 (95.8%)
Rebleed (RTO) 2 (7.7%) 0 (0%)
Mean Age 22.65 ± 3.01 22.08 ± 3.28 0.5242
Length of Hospital Stay (mean) 34 13 0.0008
Severe TBI Disability Score
Presenting GCS 3-5 1
6-8 0
Surgery Performed Craniectomy 1
Craniotomy 0
Injury Sustained Closed Head Injury 2
Depressed Skull Fracture 1
30 day percent chance of severe disability (GOS 2 and 3)
4 Points 87% *
3 Points 64%
2 Points 6%
1 Point <1%
*- includes 40% chance of Mortality and 47% chance of severe disability if patient survives 30 days
Citations:
[1] Taylor CA, Bell JM, Breiding MJ, Xu L. Traumatic Brain Injury–Related Emergency Department Visits,
Hospitalizations, and Deaths — United States, 2007 and 2013. MMWR Surveill Summ 2017;66(No. SS-
9):1–16. DOI: http://dx.doi.org/10.15585/mmwr.ss6609a1
[2] Stiver SI. Complications of decompressive craniectomy for traumatic brain injury. Neurosurg Focus.
2009;26:E7.
[3] Piek J, Chesnut RM, Marshall LF, van Berkum-Clark M, Klauber MR, Blunt BA, et al. Extracranial
complications of severe head injury. J Neurosurg. 1992;77(6):901–7.
[4]Cooper DJ, Rosenfeld JV, Murray L, Arabi YM, Davies AR, D'Urso P, et al. Decompressive craniectomy
in diffuse traumatic brain injury. N Engl J Med. 2011;364:1493–502.
[5] Zaninotto, A., Vicentini, J., Solla, D., Silva, T., Guirado, V., Feltrin, F., . . . Paiva, W. (2017). Visuospatial
memory improvement in patients with diffuse axonal injury (DAI): A 1-year follow-up study. Acta
Neuropsychiatrica, 29(1), 35-42. doi:10.1017/neu.2016.29
[6] Brooks DN. Recognition memory, and head injury. Journal of Neurology, Neurosurgery & Psychiatry
1974;37:794-801.
[7] Sophie Su YR, Veeravagu A, Grant G. Neuroplasticity after Traumatic Brain Injury. In: Laskowitz D,
Grant G, editors. Translational Research in Traumatic Brain Injury. Boca Raton (FL): CRC Press/Taylor and
Francis Group; 2016. Chapter 8. Available from: https://www.ncbi.nlm.nih.gov/books/NBK326735/
[8] Berger M.S. et al. Outcomes from severe head injury in children and adolescents. J Neurosurg.
1985;62(2):194–199
[9] Nguyen H.S., Janich K, Sharma A, Patel M, Muller W. To retain or remove the bone flap during
evacuation of acute subdural hematoma: Factors Associated with preoperative brain edema. World
Neurosugery 95: 85-90. November 2016.