Keywords:
decompressive craniectomy - brain injuries - cost and cost analysis - intracranial
pressure
Palavras-chave:
craniectomia descompressiva - lesões encefálicas - custos e análise de custo - pressão
intracraniana
INTRODUCTION
Decompressive craniectomy is a common neurosurgical procedure in which a skullcap
portion of bone is removed and an opening is made in the underlying dura mater[1], aiming to relieve raised intracranial pressure. Because of the skull rigidity,
brain swelling or focal lesions lead to increased intracranial pressure and so decrease
cerebral perfusion pressure, cerebral blood flow and oxygenation[2]. As well, with these brain hemodynamic changes, cerebral edema increases and could
finally evolve into herniation and death[3].
This procedure is used in several conditions in which this cascade occurs, such as
in stroke and in traumatic brain injury (TBI) with cerebral edema[4]. Decompressive craniectomy is a method capable of disrupting this cycle and restoring
brain functionality[1] when it is properly indicated and performed within a few hours of evaluation.
Despite several studies having described this topic, the indication for decompressive
craniectomy remains under discussion, as well as the appropriate time within which
this surgery could bring about a benefit. The effect of this procedure on long-term
functional prognosis and costs is also a concern[5]. Decompressive craniectomy has been used for raised intracranial pressure from several
etiologies, such as aneurysmal subarachnoid hemorrhage, encephalitis, acute demyelinated
encephalomyelitis, cerebral venous thrombosis and parenchymal hemorrhagic stroke.
However, the most important studies and evidences are greater for its use in TBI and
stroke[1].
Because TBI is a significant cause of morbidity and mortality globally, accurate research
on this theme is very relevant for the public health system. However, most reported
studies have described technical features, prognosis and life quality[6],[7], whereas the direct costs and economic evaluation of TBI are variables that have
been less studied but nonetheless are essential for resource allocation and improvement
of patient management.
An example of the economic impact of this situation was demonstrated by 2012 data
from the Brazilian public health system, which estimated that, in that year, about
one million patients were admitted in hospitals because of external causes (TBI included)
and the mean hospitalization cost was R$ 1,079.60 (US$ 337.38). This represented more
than one billion Brazilian Reais for the public health system's account[8].
Severe TBI presents a poor outcome in most cases[9]. Aggressive and expensive treatment for severe TBI is therefore a controversial
topic because, besides being an optimal therapeutic, it has a high morbidity and mortality
rate[5]. Prevention is essential to decrease the number of trauma patients and should be
encouraged as therapeutic approaches are very limited in improving outcomes in severe
cases. Controlling traffic speed limits, alcohol ingestion and seatbelt use help to
reduce traffic accidents that are a major cause of severe TBI.
This study aimed to evaluate the direct costs of decompressive craniectomy in patients
with TBI in a developing country and to describe the epidemiological profile of these
patients.
METHODS
An observational and retrospective study was performed using the Neurosurgical Department
database of Santa Casa de São Paulo, Brazil in the period between February 2010 and
February 2015.
All study procedures were in accordance with the ethical standards of the institutional
and/or national research committee and with the 1964 Helsinki declaration and its
later amendments or comparable ethical standards. The local institution's Human Ethics
Research Committee had previously approved this project.
An extensive literature review using the Pubmed database was also performed, with
the terms: decompressive craniectomy, brain trauma injury, costs. All patients with
doubtful and conflicting information were excluded.
All the patients who had undergone decompressive craniectomy after TBI had cerebral
hemisphere edema (with or without focal lesions, especially subdural hematoma) with
a major midline shift (greater than 5mm). A standard decompressive hemicraniectomy
was performed through a question mark or a T- (Becker) incision. A craniectomy, including
the removal of the temporal tip, was performed, followed by the dural opening. A dura
replacement was not used. In some cases, the pericranium was sutured to the dura and,
in others, a rapid closure technique was performed.
Cost calculation of the entire procedure was determined through a detailed evaluation
of surgical material, human labor, drugs, blood products, parenteral nutrition, imaging
examinations, laboratory tests, days of hospitalization and deaths. The value was
initially calculated in Brazilian Reais and later converted to US Dollars (this conversion
was performed using the January 13, 2017 exchange rate in which US$ 1 corresponds
to R$ 3.2).
After the data collection, statistical analysis was performed using a free R software
program (3.3.0 version) and IBM SPSS 20 version software. The following parameters
were used: mean, median, standard deviation, minimum and maximum for summarizing the
quantitative variables. For categorical variables, absolute and relative frequencies
(%) were applied. Fisher's exact test was used to evaluate the association among categorical
variables. For comparing two groups considering quantitative variables, the unpaired
Mann-Whitney U-test was used and for comparing more than two groups, the non-parametric
Kruskal-Wallis test was chosen.
Spearman's correlation coefficient was used to evaluate the linear correlation between
two quantitative variables. The Kaplan-Meier survival index was used for estimating
the hospitalization time until discharge. In addition, the log-rank test was performed
for comparing time until discharge.
RESULTS
During this period, the neurosurgical department performed 5,223 surgeries, of which
1,030 were TBI and, when considering only the decompressive craniectomies for trauma,
a total of 133 surgeries were performed ([Table 1]). A random selection of 32 patients with a TBI diagnosis, who underwent decompressive
craniectomy, were selected.
Table 1
Description of all neurosurgical procedures performed during the period of February
2010 to February 2015 at Santa Casa de São Paulo, Brazil.
|
Year
|
Number of surgeries
|
Traumatic brain injury surgeries
|
Decompressive craniectomy due to traumatic brain injury
|
|
2010
|
687
|
107
|
30
|
|
2011
|
866
|
164
|
14
|
|
2012
|
980
|
209
|
20
|
|
2013
|
987
|
196
|
25
|
|
2014
|
847
|
182
|
15
|
|
2015
|
856
|
172
|
29
|
|
Total
|
5223
|
1030
|
133
|
The male gender predominated in this survey, totaling 84.4%, whereas only five patients
(15.6%) were female. The average age was 44 years, the youngest patient was 15 years-old
and the oldest was 75 years of age ([Table 2]).
Table 2
General characteristics of the patients.
|
Variable
|
Number of patients
|
|
Gender
|
|
Male
|
27 (84.4%)
|
|
Female
|
5 (15.6%)
|
|
Age
|
|
15–30 years
|
5 (15.0%)
|
|
31–45 years
|
13 (41.0%)
|
|
> 45 years
|
14 (44.0%)
|
|
Admission Glasgow Coma Scale scores
|
|
Severe (≤ 8)
|
21 (65.6%)
|
|
Moderate (9–12)
|
7 (21.9%)
|
|
Mild (13–15)
|
4 (12.5%)
|
|
Time until the surgery
|
|
< 6h
|
9 (28.0%)
|
|
6–12h
|
12 (37.0%)
|
|
13–24h
|
5 (16.0%)
|
|
> 24h
|
6 (19.0%)
|
|
Operated side
|
|
Right
|
11 (34.4%)
|
|
Left
|
20 (62.5%)
|
|
Both
|
1 (3.1%)
|
|
Discharge Glasgow Outcome Score (GOS)
|
|
Mortality (GOS = 1)
|
22 (68.8%)
|
|
Poor outcome (GOS = 2 and 3)
|
5 (15.6%)
|
|
Good outcome (GOS = 4 and 5)
|
5 (15.6%)
|
Of the comorbidities, 22.0% of patients were alcoholics, 18.7% had hypertension and
6.2% had diabetes. No other associated pathologies were described.
Most of the patients had multiple brain lesions and the most frequent were acute subdural
hematoma (68.7%), diffuse brain edema (56.3%) and contusion (50.0%). Besides these,
at a lower frequency, traumatic subarachnoid hemorrhage (18.7%), epidural hematoma
(9.4%) and cranial gunshot wounds (9.4%) were described. Considering the topography,
the majority presented with a left hemisphere lesion (62.5%).
Regarding infectious complications, 69.0% developed some type of infection, such as
pneumonia, meningitis, urinary tract infection and sepsis.
Most of the patients arrived at hospital in a severe general status. On admission,
21 of the 32 (65.6%) patients presented with a Glasgow Coma Scale (GCS) ≤ 8; 21.9%
with a GCS varying between 9 and 13; and only 12.5% had mild TBI. The mean time between
arrival at the hospital and the surgical procedure was eight hours; however, with
great variability (standard deviation (SD) of 39.2 hours and variation of 2–144 hours).
The mean hospitalization time was 40.3 days with a SD of 48.9 days and variation of
2–230 days.
The economic impact of decompressive craniectomy for TBI was measured by calculating
the whole cost. A linear and positive correlation between days of hospitalization
and the cost can be seen in [Figure 1]. Considering the other variables in relation to the cost, after a linear regression
model analysis, none demonstrated a significant correlation.
Figure 1 Positive and linear relation between days of hospitalization and cost (Spearman Correlation
coefficient of 0.925)
Therefore, knowing what affects the cost, an equation was proposed considering the
surgical procedure cost and days of hospitalization for calculating of the whole cost.
The expected cost formula in Brazilian Reais (R$) is below:
* Time of hospitalization in days; R$ 3,858.921 refers to the inherent procedure cost.
For example, a patient hospitalized for 60 days has an estimated cost of R$ 92,435.35
(US$ 28,886.05) ([Figure 2]). This is a high value considering that the minimum Brazilian wage is US$ 292.81
per month (this conversion was performed using the January 13, 2017 quotation in which
US$ 1 corresponded to R$ 3.2).
Figure 2 Linear model of the cost in Brazilian reals.
The variables correlating with a greater time of hospitalization were the severity
of the trauma – patients with severe TBI had a positive correlation with increased
costs ([Figure 3]) and infectious complications led to longer hospitalization times ([Figure 4]).
Figure 3 Proportion of patients for each hospitalization time according to TBI severity.
Figure 4 Proportion of patients for each hospitalization time according to the presence of
infectious complications.
Comparing hospitalization days with the outcomes, it can be seen that survivors had
a longer period of hospitalization ([Figure 5])
Figure 5 Proportion of patients for each hospitalization time according to outcome.
Considering outcomes and costs, the general mortality index was 68.7%. Only 10 patients
survived and among them five (15.6% of the initial group of 32) had a good outcome
(Glasgow Outcome Scale (GOS) score of 4 or 5). The total cost was R$ 2,116,960.22
(US$ 661,550.06) and the mean patient cost was R$ 66,155.00 (US$ 20,673.44) (SD ±
R$ 66,766.10 (US$ 22,255.37)), and the minimum value was R$ 6,883.84 (US$ 2,151.20)
and the maximum of R$ 309,119.80 (US$ 96,599.94).
DISCUSSION
Decompressive craniectomy is a common neurosurgical procedure that aims to relieve
raised intracranial pressure. Several situations could require this surgery for restoring
encephalic hemodynamics[1]. TBI could require this procedure and, because of its global importance and socioeconomic
impact, mainly in developing countries, knowing the costs and outcomes makes a difference
in the public health system.
Removal of the skullcap, as a type of therapeutic approach, has been used since 10,000
BC[10]. Hippocrates initially described the types of brain trauma injuries that could be
treated with burr holes[11]. However, only in 1901, did Theodor Kocher establish the theory that a skull opening
could relieve raised intracranial pressure[12]. Some years later, in 1908, Harvey Cushing described his own statistics of several
patients with TBI treated with subtemporal decompressive craniectomy[13]. In the following decades, the use of decompressive craniectomy declined and few
studies were undertaken to improve the technique[12]. However, in the 1990s, with the development of neurological intensive care and
neuroimaging advances, a more extensive study of decompressive craniectomy returned
and randomized analyses were performed to evaluate its effectiveness[13]. Therefore, from this moment on, guidelines describe decompressive craniectomy for
strokes and TBI as an important and systematic therapeutic approach[12].
In the treatment of raised intracranial pressure in TBI, decompressive craniectomy
could be performed as a primary or secondary therapy. The former is performed during
surgeries for hematomas (usually acute subdural hematoma) due to cerebral edema associated
with transcalvarial herniation visualized intraoperatively, or for edema that may
get worse in the postoperative period[14],[15],[16]. On the other hand, secondary therapies are those performed as a rescue measure,
when all nonsurgical treatments for raised intracranial pressure control have not
succeeded[17],[18].
In fact, the most common technique for unilateral decompressive craniectomy is a fronto-temporo-parietal
incision with removal of the bone flap and dural opening (usually with a “starry”
incision)[19],[20]. This technique was used in the present study. Most of these patients were in poor
general condition and, therefore, neurological intensive care is very important for
improving prognosis and increased survival.
The technique and indications have already been discussed extensively in several articles[13],[20],[21]. Nevertheless, this study has a different focus and uncommon approach, namely, the
cost of decompressive craniectomy, a procedure that represents the most severe and
expensive treatment option for patients with TBI. The main indications for decompressive
hemicraniectomy are edema after spontaneous subarachnoid hemorrhage, malignant middle
cerebral artery infarct and post-traumatic edema related, or not, to adjacent focal
lesions.
Several variables were determined and compared to the patient outcome (evaluated through
the GOS). They included gender, age, type of brain lesion, associated pathologies,
the operated hemisphere side, infectious complications and admission GCS. However,
none of these had a significant correlation with the patient outcome. Another cost
analysis study involving 54 patients also did not show an association between age
and GCS admission with the patient's outcome[22], despite these variables typically being considered as prognostic factors[21].
When the time until surgery was considered, there have been some divergent studies
with conflicting data concerning outcomes. Therefore, despite this study not indicating
a significant association between these variables, others have shown better outcomes
with early surgery[3],[23] and some have also described a decrease in mortality[22],[24]. In this context, it is valuable to note that our mortality rate was 68.7%, a value
greater than many other studies, in which the mortality index was 28–41%[22],[25] with aggressive treatment, and about 41–51% with nonsurgical treatments (such as
hypothermia and hyperosmolar solutions)[9],[25]. Our high mortality rate could be because this research was performed in a large
center that receives patients from an extensive geographic area, many of whom arrive
several hours after the injury, in a non-optimal time and, therefore, in a severe
health status, contributing to increased mortality.
Thus, hospitalization days have a direct impact on TBI decompressive craniectomy patient
costs and, therefore, an analysis was performed to describe which factors are correlated
with days of hospitalization and have an influence over the final cost.
Infections could be correlated with greater costs because these patients require broad-spectrum
antibiotics for longer periods and, as well, several procedures such as gastrostomy
and tracheostomy may be performed in chronic patient care.
It has previously been described in some studies that an increase in antibiotic use
in hospitalized patients, mainly broad-spectrum, generally have higher costs[26]. In most cases, multidrug resistant bacteria were isolated. Therefore, these drugs
have significant impact on costs during hospitalization[27],[28].
Comparing hospitalization days with the outcomes, it can be seen that survivors had
a longer period of hospitalization ([Figure 5]). A large trial published in 2011, known as DECRA[17], revealed that patients undergoing craniectomy had worse scores on the Extended
GOS than those receiving standard care. As well, in a randomized study in 2016, Hutchinson
et al. observed that decompressive craniectomy could reduce mortality; however, several
of the survivors were in a vegetative state with a low functional status[18].
This study has some limitations inherent in its model. Because it was a retrospective
analysis, data was obtained through medical records and patients could not be followed
up. Another limitation is the absence of a controlled group. However, in this context,
conservative treatment for critical neurological patients is a difficult choice, considering
that a neurosurgical procedure[17] is a rescue measure and usually indicated when no other therapy has succeeded. A
nonsurgical approach is seen as an ethical dilemma.
Only the direct costs of decompressive craniectomy were described. Rehabilitation,
homecare support, social security and loss of productivity were considered as indirect
costs and not evaluated in this study.
Patients who do survive generate greater hospitalization costs and the majority will
require future hospitalizations for treatment of clinical complications and other
procedures, such as ventricular shunts and cranioplasty. Some studies have demonstrated
that the real cost of these patients is about 10-15 times greater than the cost of
the treatment in acute phase[29],[30].
This was the first Latin-American study aimed at determining the direct cost of decompressive
craniectomy in the public health system and the main factors related to the cost of
this procedure. An epidemiological profile was also described, which helps to guide
educational strategies for reduction and prevention of this socioeconomic impact.
New techniques have recently been studied to replace the decompressive craniectomy
in TBI patients. Cherian et al.[31] described a patient series in which he defends the use of microsurgical cisternostomy
in TBI patients who had a classic indication for decompressive craniectomy, revealing
a better outcome. This was a pioneer study and more research is required to validate
this technique.
In conclusion, decompressive craniectomy is a procedure that aims to relieve raised
intracranial pressure in seriously injured patients; however, it has a high mortality
rate (68%) and poor outcomes (15%). Despite the high costs in developed countries,
this was the first study that aimed to perform an analysis in developing countries.
The mean cost per patient was about R$66,000 (US$20,625) during hospitalization, with
infectious complications having a greater impact on the total cost.
The knowledge of the costs of decompressive craniectomy in developing countries in
which TBIs are very common and have a huge socioeconomic impact is important to guide
public health policies and resource allocation. Based on this study, results that
determine the high costs associated with decompressive craniectomy for TBI, lead us
to believe that prevention policies must be improved and other therapeutic modalities
should be considered in this situation.
